News Analysts, Reporters, and Journalists

27-3023.00
Median wage $62,200/yr39,250 employed (US)Rank #260 of 923 scored · top 28% by substitution

Narrate or write news stories, reviews, or commentary for print, broadcast, or other communications media such as newspapers, magazines, radio, or television. May collect and analyze information through interview, investigation, or observation.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution35
Exposure30
Augmentation67

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

30 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

7%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%29

panel mean rating 2.2/5 → substitution pressure 29/100

Technical feasibility todayw 20%30

panel mean rating 2.2/5 → substitution pressure 30/100

Cost vs. human wagew 15%37

panel mean rating 2.5/5 → substitution pressure 37/100

Adoption barriersw 20%inverted — strong barriers lower the score49

panel mean rating 3.0/5 (barrier strength) → substitution pressure 49/100

Sector adoption velocityw 10%37

panel mean rating 2.5/5 → substitution pressure 37/100

Task breakdown (30 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Check reference materials, such as books, news files, or public records, to obtain relevant facts.

74

CI 72–75 · exposure 75 · augmentation 100 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5News organizations are adopting automated research tools and AI-assisted fact-checking, but uptake remains uneven. Larger media companies and wire services use these systems; smaller outlets lag. Production deployment is growing but not yet universal.
Sector adoption velocityclaude-sonnet-54/5News and media organizations have rapidly adopted AI search and research tools, with several major outlets deploying AI-assisted research pipelines in production.
Augmentation potentialclaude-haiku-4-5-202510015/5AI substantially augments journalists and reporters by rapidly surfacing relevant facts, cross-referencing sources, and flagging related stories, allowing humans to focus on synthesis, investigation, and editorial judgment while staying fully in control of fact selection and narrative.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up locating and summarizing relevant facts from large volumes of reference material, letting journalists focus on verification and writing.
Task automatabilityclaude-haiku-4-5-202510014/5Checking reference materials and retrieving relevant facts is largely automatable today. AI systems can efficiently search databases, archives, and public records to extract factual information, and large language models can filter and organize findings, though human judgment on relevance often remains necessary for complex stories.
Task automatabilityclaude-sonnet-54/5AI can search, retrieve, and synthesize facts from digital text corpora, news archives, and public records far faster than manual lookup, meeting the time-saving bar for most routine fact-gathering.But verification of accuracy still requires human judgment, especially for ambiguous or contested facts.
Adoption barriersclaude-haiku-4-5-202510012/5Few regulatory or legal barriers prevent automated fact-checking and reference lookup. Organizational adoption is hindered mainly by institutional preference for human verification and integration friction rather than hard licensing requirements.
Adoption barriersclaude-sonnet-52/5No licensing requirement for fact-checking itself, but journalistic standards and liability for factual errors create some organizational caution around fully automated sourcing.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-powered research and retrieval tools cost a fraction of hiring a research assistant or librarian to manually check sources. API calls and inference are orders of magnitude cheaper than the fully-loaded wage for this clerical research work.
Cost vs. human wageclaude-sonnet-54/5AI-assisted search and summarization of reference materials is far cheaper per query than a reporter manually combing archives, though some oversight cost remains.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple deployed products (legal research tools, fact-checking systems, automated database search, and LLM-based retrieval systems) perform this task reliably at scale. Some edge cases and nuance-checking still benefit from human oversight, but the core function is production-ready.
Technical feasibility todayclaude-sonnet-54/5Deployed tools (search-augmented LLMs, retrieval systems, news databases with AI search) are widely used in newsrooms today for fact lookup, though hallucination risk means journalists still verify outputs.

Review written, audio, or video copy, and correct errors in content, grammar, or punctuation, following prescribed editorial style and formatting guidelines.

71

CI 66–75 · exposure 67 · augmentation 100 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Digital media and newsroom automation are mature sectors with strong AI adoption incentives. Many outlets already use AI-assisted editing tools; cost pressure on journalism accelerates uptake. Adoption is measurably growing in information-sector workflows.
Sector adoption velocityclaude-sonnet-54/5Media/publishing is a fast-digitizing information-sector industry where AI writing/editing assistants have been adopted rapidly across newsrooms and content platforms.
Augmentation potentialclaude-haiku-4-5-202510015/5AI copy-editing tools dramatically raise reporter and editor productivity by automating routine checks, freeing human judgment for complex editorial decisions. Real newsrooms report faster turnaround and improved consistency with human editors supervising the AI output.
Augmentation potentialclaude-sonnet-55/5AI dramatically speeds up copyediting workflows, flagging errors and suggesting fixes while editors retain final judgment, a clear productivity transformation with human in the loop.
Task automatabilityclaude-haiku-4-5-202510013/5AI can catch many grammar, punctuation, and formatting errors reliably, and some style violations, but editorial judgment about content accuracy, tone consistency, and nuanced style adherence still requires human oversight. Current tools automate perhaps 40–60% of the mechanical work, requiring human verification.
Task automatabilityclaude-sonnet-54/5Grammar, punctuation, and style-guide checking is a well-solved NLP task; AI copyediting tools handle most of this with substantial time savings, though nuanced editorial judgment on tone/content accuracy still needs human review.
Adoption barriersclaude-haiku-4-5-202510012/5No legal requirement mandates human copy-editors; most newsrooms operate on cost and speed. Editorial standards exist but are enforced internally, not by regulators. Organizational inertia and quality-assurance custom are modest barriers compared to professions with legal gatekeeping.
Adoption barriersclaude-sonnet-52/5No licensing requirement for copyediting, but newsrooms retain editorial oversight for liability, accuracy, and brand voice reasons, creating moderate organizational friction.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI editing tools cost pennies per document, while professional copy editors command $40–80+/hour loaded cost. Even accounting for human oversight, the marginal cost of AI is at least 5–10× cheaper than pure human copy-editing.
Cost vs. human wageclaude-sonnet-54/5AI copyediting tools cost a small fraction of a human editor's hourly wage for equivalent volume of text correction, though some human spot-checking remains needed.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple mature products (Grammarly, AP Stylebook tools, Hemingway Editor, newsroom-integrated solutions) demonstrate reliable copy-editing on grammar and basic style rules at scale in production newsrooms. Some content accuracy and voice consistency require human review, limiting full autonomy.
Technical feasibility todayclaude-sonnet-54/5Deployed products (Grammarly, AP-style checkers, LLM-based editors integrated into newsroom CMS tools) reliably catch grammar/style errors in production today, though full multimodal audio/video copy review is less mature.

Write online blog entries that address news developments or offer additional information, opinions, or commentary on news events.

70

CI 61–79 · exposure 62 · augmentation 88 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Major news organizations, digital publishers, and financial media are actively deploying AI writing tools for earnings reports, market analysis, and routine news summaries. Measurable production use is evident, though human journalists remain the primary content creators for investigative and feature work.
Sector adoption velocityclaude-sonnet-54/5Media and publishing is a fast-adopting digital sector, with many outlets already using AI for draft generation, summarization, and content augmentation in production workflows.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants substantially augment journalist productivity by drafting initial content, organizing facts, suggesting angles, and generating multiple versions for A/B testing. Journalists retain editorial control and credibility, while AI reduces time spent on routine structure and formatting.
Augmentation potentialclaude-sonnet-55/5AI is widely used to draft outlines, generate first drafts, suggest angles, and speed up research, substantially boosting a journalist's output while they retain editorial control and fact-checking responsibility.
Task automatabilityclaude-haiku-4-5-202510013/5Current AI can generate blog entries on news topics with reasonable quality and speed, potentially meeting the 50% time-saving threshold for straightforward commentary. However, original reporting, investigative depth, and credibility require human judgment and verification, limiting end-to-end automation of the full task.
Task automatabilityclaude-sonnet-54/5Blog entries summarizing news or offering commentary follow predictable formats and rely heavily on synthesizing publicly available information, which LLMs handle well; a human still typically edits for accuracy, voice, and liability, but most drafting time can be saved.
Adoption barriersclaude-haiku-4-5-202510012/5Few legal barriers exist for automated blog generation; newsrooms can publish algorithmic commentary without licensing restrictions. Organizational and reputational friction (credibility, byline authenticity, liability for errors) poses some resistance but is weaker than in regulated professions.
Adoption barriersclaude-sonnet-52/5There's no licensing requirement to publish commentary, but reputational and legal liability for factual errors or defamation creates some incentive for human review before publication.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference costs for generating a blog entry (a few cents per piece with oversight) are substantially cheaper than a journalist's fully loaded wage for the same output, though editorial review adds overhead. The cost advantage is clear but not quite an order of magnitude.
Cost vs. human wageclaude-sonnet-55/5Generating a draft blog post via LLM costs a fraction of a cent to a few cents in compute versus a journalist's hourly wage for the same output, even after factoring in editorial oversight time.
Technical feasibility todayclaude-haiku-4-5-202510013/5Deployed LLM tools and AI writing assistants reliably generate blog-length text on news topics, but material challenges remain: factual errors, outdated information, lack of genuine insight, and inability to verify sources. Products exist in production but require substantial human oversight and revision.
Technical feasibility todayclaude-sonnet-54/5Products like ChatGPT, Jasper, and newsroom-specific tools already generate draft blog posts and commentary at scale, though outlets still use human review for factual accuracy and legal risk, so it's not fully autonomous in production.

Review and evaluate notes taken about news events to isolate pertinent facts and details.

56

CI 54–59 · exposure 50 · augmentation 88 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Major newsrooms and digital publishers are experimenting with AI-assisted fact extraction and summarization, but adoption remains inconsistent and limited to pilot and helper roles rather than autonomous deployment. Adoption is accelerating but not yet deep.
Sector adoption velocityclaude-sonnet-53/5Newsrooms are adopting AI tools for transcription, summarization and research assistance at a moderate pace, though full trust in fact isolation remains limited and uneven across outlets.
Augmentation potentialclaude-haiku-4-5-202510014/5AI summaries, flagged claims, and structured extraction of details substantially assist journalists in processing large volumes of notes, allowing them to focus on editorial judgment and verification rather than manual review.
Augmentation potentialclaude-sonnet-55/5AI tools substantially speed up sorting through notes, transcripts, and interview recordings to surface key facts, letting journalists focus on verification and narrative framing.
Task automatabilityclaude-haiku-4-5-202510013/5AI can identify factual claims and extract key details from notes with reasonable accuracy, but isolating 'pertinent' facts requires editorial judgment about newsworthiness and context that varies by story. Current systems could automate parts of fact-checking and structuring, but journalists would still need to verify and prioritize.
Task automatabilityclaude-sonnet-53/5LLMs can summarize and extract key facts from notes/transcripts quickly, but journalistic judgment on newsworthiness, source credibility, and nuance still requires human review, limiting full automation.
Adoption barriersclaude-haiku-4-5-202510013/5There is no legal barrier to automating note review, but editorial standards, liability for inaccuracy, and organizational culture create friction. Newsrooms must retain human judgment over what is deemed 'pertinent,' limiting full substitution.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but reputational/liability risk from factual errors in news creates institutional caution and editorial sign-off requirements.
Cost vs. human wageclaude-haiku-4-5-202510014/5LLM inference and integration are very cheap compared to journalist wages. Even with overhead and oversight, AI-assisted fact isolation costs orders of magnitude less than paying a journalist to perform the same review.
Cost vs. human wageclaude-sonnet-54/5Automated summarization and fact-extraction via LLM APIs is very cheap per document compared to reporter time spent manually sifting notes.
Technical feasibility todayclaude-haiku-4-5-202510013/5Products like summarization tools and fact-checking assistants exist in deployment, but they operate at modest accuracy levels and require human oversight to avoid false positives or missing nuance. Newsrooms use AI-assisted extraction, but not as a fully autonomous step.
Technical feasibility todayclaude-sonnet-53/5AI transcription and summarization tools are deployed in newsrooms (e.g., auto-summarizers, note-organizing assistants) but are used as aids rather than trusted to independently isolate facts without editorial oversight.

Take pictures or video, and process them for inclusion in a story.

54

CI 32–75 · exposure 50 · augmentation 88 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Media and journalism sectors (especially digital-native newsrooms and wire services) have rapidly adopted AI-assisted photo/video workflows, automated tagging, and editing tools; production deployment is visible across major outlets and is accelerating.
Sector adoption velocityclaude-sonnet-53/5Newsrooms have adopted AI editing and image-enhancement tools moderately, but production-grade autonomous capture/processing pipelines remain uncommon industry-wide.
Augmentation potentialclaude-haiku-4-5-202510015/5AI dramatically augments journalist productivity in media capture and processing workflows: automated file organization, transcoding, color correction, and smart cropping allow reporters to focus on storytelling and editorial judgment rather than technical asset management.
Augmentation potentialclaude-sonnet-54/5AI substantially speeds up photo/video editing, tagging, cropping, color correction, and basic video editing, letting journalists focus more on capture and storytelling while AI handles post-processing tasks.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems can automatically capture, organize, and process media assets with minimal human intervention. Image and video editing, tagging, format conversion, and basic quality assessment are mature capabilities in tools like Adobe suite and cloud-native platforms, though human editorial judgment on what to include often remains necessary.
Task automatabilityclaude-sonnet-52/5Physically capturing photos/video at a scene requires human presence, judgment, and access; AI can assist with editing/processing but not the on-location capture itself, so the full task is not automatable end-to-end today.
Adoption barriersclaude-haiku-4-5-202510012/5Few regulatory or licensing barriers exist for automated media capture and processing itself; most friction comes from editorial standards, newsroom workflows, and preference to retain human judgment on imagery selection and authenticity—not legal prohibition.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but physical presence at newsworthy events, access credentials, and editorial trust/verification norms create meaningful friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-driven media processing (subscription to editing software plus cloud inference) costs substantially less than hiring dedicated photo/video technicians or editors for routine asset capture and processing, achieving rough 5–10× cost advantage for equivalent output at scale.
Cost vs. human wageclaude-sonnet-52/5Camera equipment, human presence at events, and editorial judgment still dominate costs; AI editing tools reduce some post-production time but don't replace the capture cost, keeping overall cost comparable to human labor.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (Adobe Firefly, Runway, DaVinci Resolve with AI features, cloud media management platforms) demonstrably perform image/video capture workflow acceleration and automated processing at scale in newsrooms and media organizations today, though final editorial curation typically requires human oversight.
Technical feasibility todayclaude-sonnet-52/5Some AI tools exist for photo/video editing and enhancement, but no deployed product autonomously captures newsworthy images in the field or reliably processes raw footage into finished story-ready assets without human direction.

Research a story's background information to provide complete and accurate information.

51

CI 41–61 · exposure 42 · augmentation 100 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Major news organizations have adopted AI research, summarization, and fact-checking tools in production for background research, and digital-first newsrooms are rapidly integrating such systems. However, adoption remains primarily assistive rather than fully automating the research function.
Sector adoption velocityclaude-sonnet-53/5Newsrooms are adopting AI research and summarization tools at a moderate pace, with many outlets piloting tools but maintaining human-led verification workflows.
Augmentation potentialclaude-haiku-4-5-202510015/5AI demonstrably enhances journalist productivity by rapidly gathering, organizing, and summarizing background material, enabling faster story development and broader source coverage. Journalists working with these tools can research more deeply and quickly while retaining editorial control and verification.
Augmentation potentialclaude-sonnet-55/5AI significantly speeds up background research, document summarization, and information gathering, greatly enhancing journalist productivity while they retain control over verification and framing.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can rapidly gather and summarize publicly available information, journalistic research requires verification of sources, detection of misinformation, contextual judgment about relevance, and often original reporting (interviews, documents, field work) that current AI cannot perform end-to-end. Significant human oversight and original research remain necessary.
Task automatabilityclaude-sonnet-53/5AI can quickly gather and synthesize background information, summarize documents, and search public records, but verifying accuracy, sourcing credibility, and pursuing original leads still requires human judgment and effort.The task is partially automatable with significant human oversight.
Adoption barriersclaude-haiku-4-5-202510013/5Journalistic ethics codes, editorial standards, and legal liability for inaccuracy create institutional friction against full automation. Publishers typically require human journalists to own the research and vouch for accuracy. Customer and audience expectation of human editorial judgment also limits pure substitution.
Adoption barriersclaude-sonnet-52/5There's no licensing requirement, but journalistic ethics, editorial standards, and liability for inaccurate reporting create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510013/5AI-powered research tools (web search, summarization, document analysis) cost substantially less per query than a journalist's hourly rate, but integration, fact-checking overhead, and the need for human verification reduce the per-task savings. Costs are roughly comparable when accounting for quality assurance.
Cost vs. human wageclaude-sonnet-54/5AI-assisted research tools can compile background information far faster and cheaper than a reporter working alone, though human verification adds some cost back.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI research assistants and fact-checking tools exist and are used in newsrooms, but they operate as aids rather than autonomous performers. Error rates in source verification and occasional hallucination of details mean human journalists must validate findings; no product reliably replaces the full research task without material human review.
Technical feasibility todayclaude-sonnet-53/5Products like AI research assistants and search-augmented LLMs are used in newsrooms for background research, but they still produce hallucinations and require fact-checking, limiting reliability at scale.

Communicate with readers, viewers, advertisers, or the general public via mail, email, or telephone.

47

CI 39–55 · exposure 38 · augmentation 63 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Many newsrooms use email filtering and basic automation, but substantive reader engagement via AI remains limited in practice. Adoption is confined mostly to routing and triage; deep automation of actual communication lags due to trust and quality concerns.
Sector adoption velocityclaude-sonnet-53/5Media and publishing sectors are moderately fast in adopting AI for audience engagement tools and CRM automation, though many newsrooms still rely on human staff for direct communication, especially with advertisers and sources.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist by drafting response templates, flagging urgent inquiries, and organizing reader feedback, raising journalist productivity in handling volume. However, the human journalist remains central to meaningful dialogue with the public.
Augmentation potentialclaude-sonnet-54/5AI substantially helps journalists draft responses, summarize reader feedback, and manage high email volumes, boosting efficiency while humans retain control over tone, accuracy, and relationship management.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can draft email responses and handle basic inquiries, but genuine two-way communication with readers—especially handling nuance, addressing concerns, and maintaining professional relationships—requires human judgment and context sensitivity that AI struggles with reliably. Significant manual oversight and human intervention remain necessary.
Task automatabilityclaude-sonnet-53/5AI can draft and triage routine correspondence (e.g., answering common reader queries, sorting emails, drafting form responses), but relationship-building with sources, advertisers, and audiences often requires personal judgment and trust that current AI cannot fully replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510012/5News organizations have reputational and liability concerns about automated responses to public inquiries, and audience trust often hinges on knowing they can reach a human. These are soft barriers—preference and risk rather than legal requirement—but they meaningfully slow adoption.
Adoption barriersclaude-sonnet-52/5No strict licensing requirement governs this communication task, though organizational preference for a human byline/voice and reputational risk from AI-generated missteps create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510013/5Basic email routing and templated responses via AI are inexpensive, but meaningful reader engagement still requires staff time. The all-in cost (including oversight and human follow-up) is roughly comparable to having junior staff handle inquiries directly.
Cost vs. human wageclaude-sonnet-53/5AI-assisted communication tools (drafting, triage) are cheaper per routine interaction, but oversight, correction, and handling of complex or sensitive communications keep overall costs closer to parity with human effort in many contexts.
Technical feasibility todayclaude-haiku-4-5-202510012/5While chatbots and email automation exist, they are typically used for routing and templated responses rather than substantive reader engagement. No mature production system reliably handles the full range of reader communication (complaints, tips, relationship-building) without human fallback.
Technical feasibility todayclaude-sonnet-53/5Chatbots, email auto-responders, and CRM-integrated AI tools are deployed in media organizations for reader engagement and advertiser communication, but they handle only routine interactions reliably, with humans needed for nuanced or high-value exchanges.

Select material most pertinent to presentation, and organize this material into appropriate formats.

46

CI 41–50 · exposure 45 · augmentation 75 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Digital-native and large media organizations are piloting AI-assisted curation and formatting, but adoption remains in early-to-middle stages. Smaller outlets and traditional print/broadcast newsrooms adopt more slowly; production-scale deployment of AI-led material selection without heavy human oversight is still rare across the industry.
Sector adoption velocityclaude-sonnet-53/5Media/publishing is a digitized, fast-moving sector experimenting heavily with AI tools, though full-scale editorial judgment automation remains at pilot stage.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments journalist productivity by rapidly sorting, tagging, and pre-organizing large volumes of raw material and drafting initial format structures, allowing reporters and editors to focus on judgment and narrative coherence. This assistive role is already demonstrable in modern newsrooms and can significantly accelerate the research-to-presentation phase.
Augmentation potentialclaude-sonnet-54/5AI tools substantially speed up research synthesis, summarization, and structuring drafts, letting journalists focus on selection and refinement, a strong augmentation case.
Task automatabilityclaude-haiku-4-5-202510013/5AI can assist with material selection and format organization through text summarization, categorization, and templating, achieving meaningful time savings on these components. However, the journalistic judgment of what is 'most pertinent' to a story's framing, audience, and editorial goals requires human editorial discretion, preventing full end-to-end automation at the required quality threshold.
Task automatabilityclaude-sonnet-53/5AI can draft outlines, summarize source material, and suggest structure for a story, but selecting the most newsworthy angle and appropriate framing requires editorial judgment that current AI does poorly without heavy human curation.'
Adoption barriersclaude-haiku-4-5-202510013/5Editorial responsibility and brand liability for story accuracy and appropriateness create moderate friction; editors must sign off on final material selection and framing. Newsroom culture values human editorial judgment, and audience expectations of human-curated journalism add organizational inertia, though no hard legal licensing requirement exists.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but liability for misinformation, editorial standards, and reputational risk create meaningful organizational friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference and integration costs for material selection and formatting are low, but mandatory editorial oversight and human review of AI-selected material to maintain accuracy and judgment substantially increase the all-in cost. For most newsroom workflows, this is comparable to or slightly cheaper than direct human curation but not dramatically lower.
Cost vs. human wageclaude-sonnet-53/5AI summarization is cheap per query, but the human oversight needed to verify accuracy, newsworthiness, and framing narrows the effective cost advantage to roughly comparable levels.
Technical feasibility todayclaude-haiku-4-5-202510013/5Products like content management systems with AI-assisted tagging, summarization tools (GPT-based), and automated layout systems exist and are used in newsrooms, but they typically require substantial human review and adjustment. Error rates in context selection and formatting consistency remain material, and narrow scope to templated content limits broader application.
Technical feasibility todayclaude-sonnet-52/5Some newsroom tools (e.g., AI-assisted drafting, summarization) exist but are used as aids rather than autonomous selectors/organizers of material in production journalism at scale.

Write commentaries, columns, or scripts, using computers.

44

CI 30–59 · exposure 38 · augmentation 88 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Major news organizations remain cautious; adoption is limited to experimental use and internal research rather than production replacement. Traditional journalism sectors digitize slowly on this automation front despite tech leverage elsewhere.
Sector adoption velocityclaude-sonnet-53/5Media and publishing are adopting generative AI for drafting and research at a moderate pace, with pilots widespread but full production reliance still limited due to trust and accuracy concerns.
Augmentation potentialclaude-haiku-4-5-202510014/5AI excels at assisting journalists with research synthesis, outlining, fact-checking, and stylistic refinement, meaningfully accelerating the drafting phase while journalists maintain creative control and editorial judgment over commentary and narrative framing.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up drafting, brainstorming angles, and editing for commentary and scripts while the writer retains final judgment and voice.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can generate text drafts, news commentary and opinion columns require original voice, domain expertise, and editorial judgment that current systems struggle with authentically. AI can assist with structure and research, but end-to-end replacement meeting quality parity falls short of the 50% time-saving threshold for professional journalism.
Task automatabilityclaude-sonnet-53/5AI can draft commentary and scripts quickly, but genuine opinion journalism requires original perspective, verified sourcing, and voice that current systems only partially replicate at equal quality without heavy human revision.
Adoption barriersclaude-haiku-4-5-202510013/5Editorial liability, brand reputation risk, and newsroom norms create friction against full automation. Regulatory pressure and reader expectations for human bylines and accountability provide moderate adoption barriers, though some outlets experiment with AI assistance.
Adoption barriersclaude-sonnet-52/5No licensing requirement exists for opinion writing, but reputational/liability concerns, editorial standards, and bylines create moderate organizational friction against pure automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The integration, fact-checking, editorial review, and rewriting overhead required to deploy AI-generated commentary approaches or exceeds the cost of a junior journalist for comparable final output quality.
Cost vs. human wageclaude-sonnet-54/5Generating draft text via LLMs costs a small fraction of a journalist's time-equivalent wage, though human review still adds cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5No mature product reliably produces publishable commentary or columns at professional standards without substantial human revision. AI writing tools exist but produce generic, fact-prone output requiring heavy editorial oversight, limiting real-world deployment in reputable newsrooms.
Technical feasibility todayclaude-sonnet-53/5AI writing tools are widely deployed for drafting and ideation, but newsrooms still require human editing for accuracy, voice, and liability, so full-scope reliable production use is narrower.

Revise work to meet editorial approval or to fit time or space requirements.

43

CI 28–59 · exposure 38 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Digital newsrooms and publishing platforms have adopted AI copyediting tools at a moderate pace (pilots and integrations in some major outlets), but human editorial review remains mandatory and widespread, so adoption is neither negligible nor transformative yet.
Sector adoption velocityclaude-sonnet-53/5Media and publishing have adopted AI drafting/editing tools moderately, with pilots and partial integration but not full-scale replacement of editorial revision workflows.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools that suggest edits, flag structural issues, and tighten copy for length constraints meaningfully assist reporters and junior editors, raising their revision speed and consistency while the human editorial judgment remains central to the approval process.
Augmentation potentialclaude-sonnet-54/5AI tools are widely used to suggest edits, condense text, and check style/length compliance, meaningfully speeding up the revision process while editors retain final control.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can detect grammar and basic structural issues, revising for editorial approval requires judgment about tone, narrative flow, fact alignment with house style, and editorial intent that AI cannot reliably execute end-to-end. Meaningful parts—copyediting, headline trim—can be partially automated, but not the full task to the 50% time-saving threshold.
Task automatabilityclaude-sonnet-53/5LLMs can trim, rephrase, and restructure text to fit length/style constraints reasonably well, but incorporating nuanced editorial feedback and maintaining voice/accuracy still requires human judgment for a large share of cases.atability.
Adoption barriersclaude-haiku-4-5-202510014/5Editorial approval is a gatekeeping function tied to byline responsibility, brand liability, and publication credibility; most outlets require human editors to legally and professionally sign off on published material. This creates a hard barrier to full automation, though AI can support the process.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but editorial accountability, brand voice, and accuracy/liability concerns mean a human typically must approve final published text.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI writing assistance tools are subscription-based and still require human editors to review and approve output, making the all-in cost (tool + oversight) only marginally cheaper than direct human revision; for enterprise editorial teams, the overhead is often comparable.
Cost vs. human wageclaude-sonnet-54/5Automated text revision via AI tools costs a small fraction of a reporter's or editor's time for mechanical trimming/formatting tasks, though human review adds some cost back.
Technical feasibility todayclaude-haiku-4-5-202510012/5Grammar checkers and AI writing assistants exist in production (Grammarly, editorial tools), but they address only surface-level revision. No deployed product reliably handles the full editorial approval workflow—understanding specific publication standards, fact-checking alignment, and narrative purpose—without human oversight.
Technical feasibility todayclaude-sonnet-53/5AI writing assistants and editing tools are deployed in newsrooms for copy-editing and length trimming, but reliable handling of substantive editorial revisions (tone, angle, fact-sensitive cuts) is still narrow and error-prone.

Write columns, editorials, commentaries, or reviews that interpret events or offer opinions.

42

CI 25–59 · exposure 38 · augmentation 75 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Journalism is undergoing rapid digitization, but newsrooms are adopting AI mainly for fact-checking, headline generation, and distribution—not for core opinion writing. Few major outlets have moved opinion columns to AI; adoption remains experimental and cautious.
Sector adoption velocityclaude-sonnet-53/5Media/publishing is a digitized, fast-adopting sector for AI tools generally, but adoption specifically for opinion/editorial writing (versus templated news) is more cautious and pilot-stage due to voice and credibility concerns.
Augmentation potentialclaude-haiku-4-5-202510014/5AI excels at assisting columnists with research synthesis, outlining arguments, drafting opening paragraphs, and checking facts—tasks that can meaningfully accelerate a human writer's workflow while the journalist retains editorial control and voice.
Augmentation potentialclaude-sonnet-54/5AI is widely useful for brainstorming angles, drafting structure, summarizing background research, and polishing prose, substantially speeding up a columnist's workflow while the human retains final judgment and voice.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can draft opinion pieces and commentaries with reasonable fluency, but editorial writing requires nuanced judgment, original argumentation, and credibility that demand human oversight. AI cannot independently achieve the 50% time savings at equal quality threshold without substantial human revision and fact-checking.
Task automatabilityclaude-sonnet-53/5AI can draft opinion pieces and reviews given a topic and stance, but genuine original analysis, insider sourcing, and distinctive voice that readers value still require significant human input, so full end-to-end substitution at equal quality is only partial.
Adoption barriersclaude-haiku-4-5-202510014/5Significant barriers exist: editorial masthead accountability, byline attribution (legal and ethical), potential liability for inaccurate or misleading opinions, and strong reader/advertiser preference for human-authored commentary. Publications face reputational risk automating opinion content wholesale.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but reputational risk, byline authenticity, editorial standards, and potential liability for defamation/opinion accuracy create real organizational friction against pure AI authorship.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference is cheap, but the integration and oversight costs are substantial because editorial content requires human editorial judgment, legal review, and reputational vetting. The all-in cost per publishable column remains comparable to or exceeds hiring a freelance columnist.
Cost vs. human wageclaude-sonnet-54/5Generating draft text via LLM inference is extremely cheap compared to a journalist's salary, though human editing/fact-checking still adds cost, keeping it just below the top tier.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI language models can generate plausible opinion content, no deployed journalistic product reliably produces publication-ready columns or editorials that match professional standards for originality, voice, and editorial integrity. Most deployments remain experimental or assistive rather than end-to-end production systems.
Technical feasibility todayclaude-sonnet-53/5AI drafting tools are used in some newsrooms for first drafts and templated commentary (e.g., earnings summaries, sports recaps), but signed opinion columns and reviews at major outlets remain human-authored with AI at most assisting behind the scenes.

Develop ideas or material for columns or commentaries by analyzing and interpreting news, current issues, or personal experiences.

41

CI 36–45 · exposure 25 · augmentation 75 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5News organizations are piloting AI for headline generation, summarization, and routine story assistance, but adoption of AI for original opinion/commentary development remains limited and cautious. Most production use is still experimental rather than mainstream displacement.
Sector adoption velocityclaude-sonnet-53/5Media/publishing is a fast-digitizing sector experimenting with AI drafting tools, though opinion/commentary specifically sees more caution than straight news due to authenticity concerns.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools can meaningfully assist journalists by rapidly synthesizing background research, generating initial drafts, suggesting angles, or organizing notes—materially speeding up ideation and drafting phases while the journalist retains editorial control and voice.
Augmentation potentialclaude-sonnet-54/5AI is widely used to research background, summarize competing viewpoints, suggest angles, and produce first drafts that writers then refine with their own perspective and voice.
Task automatabilityclaude-haiku-4-5-202510012/5AI can generate draft commentary or interpret news events, but developing original opinion pieces or columns requiring distinctive voice, editorial judgment, and deep contextual understanding remains largely human-dependent. Current systems struggle with the synthesis of personal experience and nuanced interpretation that defines quality columns.
Task automatabilityclaude-sonnet-52/5Generating opinion/commentary requires original interpretation, personal voice, and lived experience that current AI cannot authentically replicate, though it can draft boilerplate takes or summarize background.
Adoption barriersclaude-haiku-4-5-202510013/5Editorial standards, byline authenticity, and reader expectations create meaningful friction—audiences expect human judgment and voice in opinion pieces. Some outlets have policies against AI-generated opinion content, and legal/liability concerns around misattribution or false statements add oversight requirements.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but reputational/liability concerns, need for authentic voice and lived experience, and audience trust create moderate resistance to full automation.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference costs for generating draft material are very low (pennies per piece), while a journalist's time on ideation and drafting is expensive (hourly labor rate). Even with necessary human oversight, the all-in cost per usable draft is substantially below human-only composition.
Cost vs. human wageclaude-sonnet-53/5AI drafting assistance is cheap per word, but editorial oversight, fact-checking, and voice correction required to make output publishable narrows the cost advantage significantly.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI text generation tools exist and can produce draft commentary, but no production system reliably produces publication-ready columns or commentaries that meet editorial standards for originality, voice, and accuracy across diverse topics. Outputs typically require substantial human revision and fact-checking.
Technical feasibility todayclaude-sonnet-52/5AI tools can draft generic commentary or brainstorm angles, but no production system reliably produces publishable, insightful opinion journalism without heavy human rewriting.

Analyze and interpret news and information received from various sources to broadcast the information.

40

CI 30–50 · exposure 42 · augmentation 88 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Media organizations are cautious adopters; most use AI for supplementary tasks (tagging, distribution) rather than replacing core news analysis and interpretation. Pilots are common but production displacement remains limited due to liability and brand concerns.
Sector adoption velocityclaude-sonnet-53/5Media organizations are piloting AI tools for research, summarization, and drafting, but broad newsroom-wide production deployment for interpretive analysis is still uneven and cautious due to accuracy and trust concerns.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist journalists by quickly summarizing source material, flagging relevant information, and drafting initial summaries, allowing humans to focus on analysis and interpretation. Productivity gains are significant where humans maintain editorial control and final judgment.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up gathering, cross-referencing, and summarizing information from disparate sources, letting journalists focus on verification, framing, and on-air delivery.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can extract and summarize information from sources, the core task requires editorial judgment, context evaluation, and interpretation that distinguishes journalism from raw data aggregation. Current systems cannot reliably perform the full analytical and interpretive layer at 50% time savings with equal quality.
Task automatabilityclaude-sonnet-53/5AI can summarize, aggregate, and draft interpretive copy from multiple sources quickly, but genuine editorial judgment, source verification, and nuanced interpretation still require human oversight for at least half the workflow.
Adoption barriersclaude-haiku-4-5-202510014/5Strong organizational and reputational barriers exist: media outlets face legal liability for inaccurate reporting, journalistic ethics require human editorial judgment and byline accountability, and audience trust depends on perceived human editorial oversight. Regulatory expectations and brand risk create high friction.
Adoption barriersclaude-sonnet-53/5No licensing requirement exists for journalists generally, but editorial standards, defamation liability, and audience trust in named human reporters create meaningful organizational and reputational friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference for summarization is cheap, but the oversight cost—required editorial review, fact-checking, and correction—remains substantial. Human-in-the-loop operation keeps total cost comparable to or higher than simple human reporting for quality-equivalent output.
Cost vs. human wageclaude-sonnet-53/5AI can cheaply process and summarize large volumes of source material, but the need for human verification and editorial oversight to avoid reputational/legal risk keeps blended costs closer to parity rather than an order-of-magnitude savings.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI writing tools exist and some news organizations use them for fact-checking or draft assistance, but no deployed product reliably analyzes and interprets complex news across diverse sources with professional editorial standards. Error rates and narrow scope limit production deployment.
Technical feasibility todayclaude-sonnet-53/5Products like AI news aggregators and drafting assistants (e.g., newsroom copilot tools) are in use, but fact-checking, bias detection, and interpretive framing remain error-prone and are not fully trusted in production without human review.

Report news stories for publication or broadcast, describing the background and details of events.

40

CI 25–55 · exposure 38 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Newsrooms are experimenting with AI for routine summaries and data-heavy reporting (sports scores, earnings), but original investigative and breaking news reporting remains predominantly human-driven. Digitally native outlets have adopted some automation, but traditional and mid-market news organizations move slowly due to trust and quality concerns.
Sector adoption velocityclaude-sonnet-53/5News organizations have adopted AI for templated reports and drafting assistance at a moderate pace, though core reporting remains largely human-driven with pilots more common than full deployment.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists journalists by rapidly summarizing sources, identifying patterns in data, generating first-draft structures, and automating routine tasks like transcription and fact-checking. These tools measurably raise productivity and free reporters to focus on investigation, interviewing, and analysis while humans remain the final decision-maker on publication.
Augmentation potentialclaude-sonnet-54/5AI substantially assists reporters with research summarization, background gathering, transcript analysis, and draft generation, meaningfully speeding up the reporting process while humans verify and add original insight.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can draft factual summaries and synthesize information from multiple sources, but news reporting requires real-time investigation, source verification, and editorial judgment that AI cannot reliably perform end-to-end. AI struggles with on-the-ground reporting, ethical judgment calls, and meeting journalistic standards at quality parity with human reporters.
Task automatabilityclaude-sonnet-53/5AI can draft basic factual reports (e.g., earnings, sports scores, weather) from structured data, but original reporting requires sourcing, verification, and on-the-ground observation that current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Journalism has strong professional norms, editorial responsibility for accuracy and bias, and reader/audience expectations for human authorship and accountability. Regulatory frameworks, press freedom conventions, and legal liability for false reporting create meaningful friction against full automation of news reporting.
Adoption barriersclaude-sonnet-52/5No licensing requirement for journalists, but editorial standards, defamation liability, and source verification create meaningful organizational friction around fully автоматized reporting.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI text generation is cheap per token, but comprehensive reporting requires human-level investigation, fact-checking, and editorial oversight that amplifies total cost. A journalist's loaded wage often remains lower than the combined cost of AI systems plus extensive human quality assurance and liability insurance.
Cost vs. human wageclaude-sonnet-53/5For templated data reports AI is far cheaper, but for full reporting requiring interviews and fact-checking, human oversight and verification costs keep totals comparable to human labor.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some AI tools generate summaries from existing information (e.g., automated news briefs from wire feeds), but no deployed product reliably conducts independent reporting, verifies sources, or produces publication-ready news stories autonomously. Real newsrooms do not operationally replace reporters with AI for original reporting.
Technical feasibility todayclaude-sonnet-53/5Automated journalism products (AP's Automated Insights, Bloomberg's Cyborg) reliably produce templated data-driven stories in production, but general investigative or event-based reporting isn't handled by deployed products at scale.

Receive assignments or evaluate leads or tips to develop story ideas.

33

CI 30–35 · exposure 25 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While media organizations use automated news feeds and topic monitoring, adoption of AI for actual lead evaluation and assignment decisions is limited to pilots or internal experimentation; production displacement remains minimal.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are experimenting with AI for tip triage and trend spotting but adoption for actual story idea generation and lead evaluation remains in early pilot stages.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by ranking, summarizing, or organizing incoming tips and leads, helping journalists scan more material faster, but the human editor remains firmly in control of final assignment and story development decisions.
Augmentation potentialclaude-sonnet-54/5AI can efficiently scan social media, documents, and databases to surface potential leads and summarize background, meaningfully speeding up a reporter's initial research and idea development.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist in filtering, organizing, and summarizing leads or tips, but the core task of evaluating story worthiness and deciding what to pursue requires human editorial judgment, contextual awareness of audience relevance, and newsroom priorities that AI cannot reliably replicate.
Task automatabilityclaude-sonnet-52/5AI can help surface trending topics or summarize tip content, but genuinely evaluating source credibility, newsworthiness, and editorial judgment on which leads to pursue remains largely human-driven.'},'rating reflects limited but real partial automation.:not applicable field misuse - ignore this text."
Adoption barriersclaude-haiku-4-5-202510013/5Editorial judgment and story assignment remain strongly human-centered functions; newsroom culture and editorial policy typically require human editors to make final assignment calls, though no strict legal licensing barrier exists.
Adoption barriersclaude-sonnet-52/5No licensing barrier, but editorial trust, source relationships, and liability for publishing errors create organizational friction against fully automating this judgment call.
Cost vs. human wageclaude-haiku-4-5-202510012/5Existing AI tools for lead aggregation require human review and integration; the total cost of setup, oversight, and error correction remains comparable to or exceeds the cost of a journalist's assignment-evaluation time.
Cost vs. human wageclaude-sonnet-52/5AI tools for lead scanning are cheap to run but still require significant human review and judgment, so all-in cost savings versus a reporter's time are modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some products exist for news topic aggregation and lead classification (e.g., news monitoring tools), but none can end-to-end evaluate leads or develop story ideas with the editorial discernment required in production newsrooms; humans still make final assignment and selection decisions.
Technical feasibility todayclaude-sonnet-52/5Some newsroom tools use AI to flag trending stories or cluster tips, but no deployed product reliably evaluates lead quality or generates assignment-worthy story ideas at scale in production.

Determine a published or broadcasted story's emphasis, length, and format, organizing material accordingly.

32

CI 28–36 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Many news organizations use AI for scheduling, categorization, and basic formatting, but adoption of AI for autonomous emphasis and editorial framing remains limited and cautious. Pilots exist, but production deployment of AI making core editorial decisions is rare due to brand and liability concerns.
Sector adoption velocityclaude-sonnet-53/5Media/publishing is a moderately fast-adopting information sector experimenting with AI for drafting and templated content, but editorial judgment tasks remain in pilot/augmentation phase rather than full deployment.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist journalists by suggesting formatting options, organizing material into structures, and flagging word counts or structural issues, but the human editor must ultimately decide emphasis and framing. This creates genuine productivity gains on the organizational and structural aspects of the task.
Augmentation potentialclaude-sonnet-54/5AI can effectively assist by generating draft structures, suggesting length/format options, and summarizing source material, significantly speeding up a journalist's decision-making while they retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with organizing and formatting material, but determining a story's emphasis requires editorial judgment that reflects audience, newsroom values, and breaking context. Current systems cannot reliably make these nuanced decisions end-to-end without substantial human oversight, preventing the 50% time-savings threshold.
Task automatabilityclaude-sonnet-52/5This requires editorial judgment about newsworthiness, audience, and outlet priorities that current AI can approximate but not reliably replace for high-stakes or nuanced editorial decisions; some structuring can be automated but the core judgment call resists full automation.
Adoption barriersclaude-haiku-4-5-202510014/5Editorial decisions about story emphasis and format carry reputational and accuracy risk, creating strong organizational friction and liability concerns. Newsrooms have editorial standards and ethical guidelines that require human judgment, and audiences expect human editorial responsibility for framing.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but editorial responsibility, brand reputation, legal liability (defamation, accuracy) and audience trust create meaningful organizational friction against fully ceding this judgment to AI.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI systems for story organization are relatively low-cost, but integration with newsroom workflows, fact-checking oversight, and editorial review adds overhead. The cost difference is modest rather than clearly favorable, especially given the need for human sign-off on emphasis decisions.
Cost vs. human wageclaude-sonnet-53/5AI drafting/formatting assistance is cheap per use, but the human editorial oversight required to validate emphasis and framing keeps overall cost comparable rather than dramatically cheaper.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools exist for text formatting and basic organization, no deployed product reliably determines story emphasis or editorial framing without human intervention. Newsrooms use AI for copy editing and scheduling, but not for autonomous editorial decision-making at quality parity.
Technical feasibility todayclaude-sonnet-52/5AI tools can suggest headlines, lengths, or formats (e.g., news aggregation and templated content systems), but deployed products rarely make final editorial emphasis/format decisions autonomously in professional newsrooms.

Examine news items of local, national, and international significance to determine topics to address, or obtain assignments from editorial staff members.

31

CI 30–32 · exposure 25 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Newsrooms have adopted aggregation and trending tools but rarely deploy them for autonomous assignment; assignment remains a human editorial function. Adoption is limited to assistive dashboards rather than replacement systems, reflecting cultural and structural resistance in journalism.
Sector adoption velocityclaude-sonnet-53/5Media/journalism sector shows moderate AI adoption for research and monitoring tools, but assignment and topic-selection workflows are still largely human-led with pilots ongoing.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools that surface trending topics, flag emerging stories, and summarize news landscapes substantially assist human editors in scanning and prioritizing what to cover. Such systems are already in use and demonstrably improve speed of story discovery while preserving human editorial control.
Augmentation potentialclaude-sonnet-54/5AI tools like news aggregators, trend analysis, and topic clustering significantly help journalists scan and prioritize potential stories faster, while humans retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5AI can scan and categorize news items by topic and significance, but determining *what topics to address* requires editorial judgment, news judgment, audience understanding, and strategic prioritization that remains highly contextual and human-dependent. Current systems lack the judgment to autonomously decide coverage strategy.
Task automatabilityclaude-sonnet-52/5AI can surface trending topics and summarize news volume, but determining editorially significant angles and receiving human assignments requires judgment and organizational context AI cannot fully replace.ed
Adoption barriersclaude-haiku-4-5-202510013/5Editorial discretion and news judgment are traditionally human responsibilities, and most newsrooms view assignment as requiring newsroom knowledge and institutional authority that resist full automation. However, there is no hard legal barrier or licensing requirement preventing AI assistance.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but editorial accountability, legal liability for content decisions, and organizational hierarchy create real friction against full automation of assignment-setting.
Cost vs. human wageclaude-haiku-4-5-202510012/5News aggregation APIs and AI scanning systems are cheap, but the human editorial judgment they replace commands significant wage value. The cost of the AI infrastructure plus required editorial oversight approaches or exceeds the cost of a junior reporter doing initial story scanning.
Cost vs. human wageclaude-sonnet-52/5AI monitoring tools are cheap to run but still require substantial human oversight and editorial judgment layered on top, limiting net savings for this specific task.
Technical feasibility todayclaude-haiku-4-5-202510012/5News aggregation and topic extraction tools exist and can identify trending stories, but reliable end-to-end assignment and prioritization at editorial quality standards requires human oversight. No deployed product independently makes newsroom assignment decisions without human editorial validation.
Technical feasibility todayclaude-sonnet-52/5Some newsroom tools use AI for trend detection and content curation, but assignment decisions and topic selection remain human-driven in production newsrooms today.

Report on specialized fields such as medicine, green technology, environmental issues, science, politics, sports, arts, consumer affairs, business, religion, crime, or education.

31

CI 25–36 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption in newsrooms remains limited to draft assistance and data aggregation; full automation of reporting is rare. Major outlets have experimented with AI-generated summaries but retain human reporters for original investigation. Smaller outlets and wire services use more automation, but the sector overall shows cautious, slow adoption rather than deep displacement.
Sector adoption velocityclaude-sonnet-53/5News organizations are piloting AI for data-driven beats (earnings reports, sports scores) but specialized investigative and expert reporting remains largely human-driven, so adoption is uneven and moderate.
Augmentation potentialclaude-haiku-4-5-202510013/5AI assists reporters by generating research summaries, organizing specialized information, and drafting outlines on technical topics, which can speed up initial reporting phases. However, augmentation is partial—interviews, investigation, and final editorial judgment remain human-driven—making AI a useful but limited productivity tool rather than a transformative force.
Augmentation potentialclaude-sonnet-54/5AI substantially helps journalists with research synthesis, background summarization, drafting, and translation across specialized fields, meaningfully boosting productivity while human judgment and verification remain central.
Task automatabilityclaude-haiku-4-5-202510012/5AI can draft initial summaries and structure basic facts on specialized topics, but journalism requires original investigation, source verification, critical analysis, and judgment about newsworthiness that AI cannot reliably perform end-to-end. The 50% time-saving threshold is not consistently met because human reporting work is dominated by activities like interviews, fact-checking, and editorial synthesis that remain primarily human-driven.
Task automatabilityclaude-sonnet-52/5AI can draft summaries or background from existing data, but original specialized reporting requires source cultivation, verification, and field expertise that current systems cannot reliably perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Significant barriers exist: editorial responsibility and reputational liability fall on the news organization (not the AI), publishers are legally accountable for accuracy and defamation, audience trust depends on human editorial judgment, and most major outlets maintain policies requiring human verification. These create organizational and legal friction that limits substitution.
Adoption barriersclaude-sonnet-53/5No licensing requirement for journalists generally, but liability for factual errors, defamation risk, and audience trust in bylined human expertise create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Cost-per-story from AI remains comparable to or higher than a junior reporter when accounting for integration, fact-checking overhead, legal review, and the human editor time needed to validate outputs. The inference cost alone is low, but the total system cost (human oversight, corrections, liability insurance) negates cost advantage.
Cost vs. human wageclaude-sonnet-53/5AI drafting is cheap for templated recaps, but accurate specialized reporting still needs costly human verification, interviews, and editorial oversight, narrowing the cost advantage.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate text summaries of specialized topics (via tools like GPT), no deployed product reliably performs complete news reporting independently. Existing systems lack the investigative capability, source verification, and editorial judgment required; they are used as draft aids, not production replacements. Products that attempt automated journalism produce material errors and miss story context.
Technical feasibility todayclaude-sonnet-52/5AI writing tools exist and are used for basic financial/sports data recaps, but no deployed product independently reports on specialized fields requiring original investigation or expert interviews at scale.

Gather information and develop perspectives about news subjects through research, interviews, observation, and experience.

29

CI 25–32 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption remains slow. Most news organizations use AI for routine tasks (wire story formatting, tagging), but core reporting and perspective-development remain heavily human-driven. Pilots exist but production displacement is minimal, reflecting cultural and legal friction in journalism.
Sector adoption velocityclaude-sonnet-53/5Media organizations are piloting AI for research assistance and drafting, but original reporting workflows remain largely human-driven with cautious, uneven adoption across the industry.
Augmentation potentialclaude-haiku-4-5-202510013/5AI assists meaningfully with research acceleration (summarizing documents, aggregating sources) and fact-checking, raising reporter productivity on information-gathering phases. However, augmentation is bounded by the need for human judgment in interviews, source evaluation, and perspective synthesis.
Augmentation potentialclaude-sonnet-54/5AI tools substantially speed up background research, source document review, transcription, and initial synthesis, meaningfully boosting journalist productivity even though the core original-reporting work remains human-led.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with information gathering via web search and data aggregation, developing perspectives requires contextual judgment, source credibility assessment, and nuanced understanding that current systems cannot reliably perform end-to-end. The task fundamentally depends on human interpretation and professional editorial judgment.
Task automatabilityclaude-sonnet-52/5AI can assist with background research and summarization, but original interviewing, on-the-ground observation, and developing a distinctive editorial perspective require human judgment, access, and credibility that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Significant barriers exist: editorial liability for inaccuracy, audience trust requirements for human-bylined reporting, regulatory expectations around journalistic ethics, and organizational preference for human reporters who take professional responsibility for their work. Many outlets legally require human editorial review.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but journalistic ethics, source trust, legal liability for defamation/accuracy, and outlet editorial standards create meaningful friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI research tools (web search, summarization) are cheap, but the bottleneck is the human journalist's time spent interpreting, interviewing, and synthesizing—the core value-add. AI does not yet replace this labor cost sufficiently to create a favorable cost ratio.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply summarize existing text, but the core value-add of original reporting (interviews, fieldwork, verification) still requires paid human labor, so overall cost savings are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product performs this task reliably as a complete workflow. AI tools can support research and fact-checking, but perspective development and conducting interviews remain human-dependent activities; deployed systems lack the judgment and contextual understanding needed for production journalism.
Technical feasibility todayclaude-sonnet-52/5Products exist for research aggregation and transcript summarization, but no deployed system reliably conducts original interviews, builds source relationships, or forms independent editorial perspectives in production newsrooms.

Present news stories, and introduce in-depth videotaped segments or live transmissions from on-the-scene reporters.

26

CI 23–30 · exposure 25 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite decades of technology development, adoption of AI news presenters remains negligible in mainstream broadcasting; most pilots (e.g., Yomou in China, limited experiments elsewhere) serve niche or cost-constrained outlets and have not scaled, reflecting slow movement in traditional broadcast sectors.
Sector adoption velocityclaude-sonnet-52/5Broadcast journalism is adopting AI for research, transcription, and content generation but on-air anchoring remains a rare novelty rather than a producitonalized trend.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist journalists by generating script suggestions, summarizing raw footage, or organizing video segments, but the core task of on-air presentation fundamentally requires the human broadcaster, limiting meaningful augmentation without full replacement (which remains infeasible).
Augmentation potentialclaude-sonnet-53/5AI can assist with teleprompter scripting, story summarization, and cueing but does not transform the live on-air presentation task itself.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can generate text summaries and compile video segments, the task of presenting news stories and introducing live segments requires real-time judgment, vocal delivery, emotional authenticity, and the ability to respond to breaking developments—capabilities current AI systems cannot reliably replicate at broadcast quality without substantial human oversight and intervention.
Task automatabilityclaude-sonnet-52/5On-camera presence, live delivery, and real-time handling of unscripted transitions to field reporters require physical embodiment and real-time judgment that current AI cannot fully replicate end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Significant barriers exist: audience trust and preference for human presenters, regulatory scrutiny of synthetic media, broadcaster liability for misinformation, FCC and similar regulatory oversight of broadcast content, and institutional resistance to automation of the on-air face of news organizations.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for anchors, but audience trust, brand identity, and live editorial judgment create substantial organizational and reputational friction against replacement.
Cost vs. human wageclaude-haiku-4-5-202510012/5The integration, training, oversight, and liability costs of deploying an AI news anchor, plus the need to maintain human backup and verify outputs, remain comparable to or exceed the loaded cost of a professional human presenter in most newsrooms.
Cost vs. human wageclaude-sonnet-52/5Building and maintaining a realistic AI avatar with live integration to field feeds requires significant technical infrastructure, offsetting savings versus a human anchor's wage in most markets.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI video synthesis and text-to-speech exist in research and limited pilots, but no mature production system reliably performs live news presentation with the fluency, credibility, and error tolerance required by broadcasting standards; deepfakes and synthetic presenters remain rare in mainstream newsrooms due to quality and trust concerns.
Technical feasibility todayclaude-sonnet-52/5AI news anchors exist in limited pilot deployments (e.g., some Asian broadcasters) but are not widely adopted in mainstream production newsrooms for live, unscripted anchoring.

Photograph or videotape news events.

26

CI 21–30 · exposure 20 · augmentation 50 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of automated video systems in news is slow; most newsrooms still deploy human videographers and photographers. Digital-native and large-scale news organizations experiment with drones and edited clips, but core breaking-news capture remains human-centric.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for editing, transcription, and drafting, but the physical act of shooting footage in the field is minimally touched by current adoption trends.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist photographers through auto-stabilization, post-production editing, object detection, and shot organization, meaningfully speeding up workflow. However, the human remains essential for framing, composition, and deciding what matters journalistically.
Augmentation potentialclaude-sonnet-53/5AI can assist with drone piloting, smart camera settings, real-time footage tagging, and post-production, improving efficiency around the human's core filming task.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI cannot autonomously decide what to photograph, frame shots, or respond to unpredictable news events in real-time. While AI can edit, stabilize, or process footage post-capture, the core decision-making and live capture of newsworthy moments remains human-dependent.
Task automatabilityclaude-sonnet-52/5Physical capture of news events requires being on-site with a camera, judgment about framing, and reacting to unpredictable live scenes; current AI cannot replace physical presence and real-time visual judgment.
Adoption barriersclaude-haiku-4-5-202510013/5Editorial judgment, safety, and liability around automated capture are material friction points. News organizations prefer human reporters for authentication, composition, and ethical framing; there are no hard legal barriers, but organizational trust in human-captured content persists.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but physical presence, access credentials, safety, and journalistic ethics/verification norms create real-world friction against remote/AI substitution.
Cost vs. human wageclaude-haiku-4-5-202510012/5High-quality camera equipment, drones, and autonomous systems are expensive; integrating them into newsrooms with oversight adds cost. A human photographer's wages remain competitive when accounting for equipment amortization, maintenance, and the need for human judgment and safety.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for on-location photography/videography, so no meaningful cost comparison favors AI; a human with equipment is required.
Technical feasibility todayclaude-haiku-4-5-202510012/5Stationary security cameras and simple automated video systems exist, but no mainstream deployed product reliably captures breaking news events autonomously with journalistic judgment. Robotics and autonomous drones show promise but lack the adaptive intelligence needed for real news environments.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously travels to and films breaking news events end-to-end; this remains a physical, human-executed task.

Assign stories to other reporters or duties to production staff.

25

CI 20–30 · exposure 20 · augmentation 50 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While newsrooms are digitizing workflows, assignment decisions remain predominantly manual and editorial, with little public evidence of AI agents displacing assignment duties even in digitally mature organizations. Adoption has been limited to scheduling tools rather than autonomous decision-making.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for drafting, transcription, and research, but assignment/editorial management functions show minimal automation deployment industry-wide.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by surfacing reporter availability, past beat coverage, and story complexity to inform an editor's decision, raising speed and consistency of the assignment process without removing human judgment from the final call.
Augmentation potentialclaude-sonnet-53/5AI can assist editors by summarizing reporter workloads, suggesting story leads, or tracking beats, providing moderate support while the human retains assignment authority.
Task automatabilityclaude-haiku-4-5-202510012/5Assigning stories and duties requires understanding reporter expertise, current workload, story complexity, and organizational priorities—nuanced judgment that AI struggles with reliably. While AI could suggest assignments based on beat and availability, final decisions involve contextual knowledge and human accountability that prevents >50% time savings at equal quality.
Task automatabilityclaude-sonnet-52/5Assigning stories requires editorial judgment about newsworthiness, reporter strengths, and workload balancing that current AI cannot reliably replicate end-to-end, though scheduling/logistics sub-parts could be partially automated.'
Adoption barriersclaude-haiku-4-5-202510014/5Editors and newsroom managers have institutional authority and responsibility for staff assignments; there is strong organizational friction and human-contact requirement, as reporter morale, fairness, and editorial judgment are expected to remain human-driven. Union agreements in some outlets may further restrict automation.
Adoption barriersclaude-sonnet-53/5No licensing requirement exists, but strong organizational and editorial-judgment norms, plus accountability for content decisions, create moderate friction against full automation of assignment decisions.
Cost vs. human wageclaude-haiku-4-5-202510012/5The integrated cost of AI assignment systems (setup, maintenance, oversight to catch errors) remains comparable to or higher than the time a senior editor spends making these decisions, especially given the cost of assignment mistakes.
Cost vs. human wageclaude-sonnet-52/5Since no reliable AI product performs this task, any attempted automation would require heavy human oversight and correction, making the effective cost comparable to or higher than a human editor's time.
Technical feasibility todayclaude-haiku-4-5-202510012/5No production systems reliably handle news desk assignment decisions autonomously; the task requires organizational context, reporter skill matching, and editorial judgment that exceed current AI capabilities. Systems might exist to flag workload or suggest templates, but humans retain substantive control.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs editorial story assignment in newsrooms today; this remains a human managerial function with no commercial AI substitute in production use.

Transmit news stories or reporting information from remote locations, using equipment such as satellite phones, telephones, fax machines, or modems.

23

CI 11–35 · exposure 13 · augmentation 38 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5News organizations have adopted automated data pipelines for structured content, but remote field transmission still relies heavily on human journalists managing equipment and connectivity decisions in unpredictable conditions. Adoption of full automation remains slow due to the unpredictability of field environments and editorial control requirements.
Sector adoption velocityclaude-sonnet-52/5Field journalism in remote/hostile environments remains low-digitization and equipment-dependent, showing slow AI integration for the physical transmission aspect specifically.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by pre-staging transmission formats, suggesting optimal transmission protocols based on available bandwidth, managing backup queues, or automating acknowledgment logging. These augmentations provide moderate productivity benefit while the journalist retains control over when and how information is transmitted.
Augmentation potentialclaude-sonnet-52/5AI can assist with drafting or formatting content before transmission, but offers little help with the actual technical act of transmitting via specialized remote equipment.
Task automatabilityclaude-haiku-4-5-202510011/5This task is fundamentally about physically transmitting already-reported information via specific hardware devices from remote locations. AI cannot independently choose transmission methods, operate physical equipment in field settings, or manage the logistics of remote connectivity in real time.
Task automatabilityclaude-sonnet-52/5The core action is operating communication equipment to send content, which is mostly a manual/technical logistics task not well-suited to AI automation; AI doesn't operate satellite phones or fax machines physically or make transmission decisions in the field.
Adoption barriersclaude-haiku-4-5-202510014/5News organizations retain editorial control and accountability for what is transmitted, creating organizational and liability barriers. Additionally, the task involves real-time field decision-making where human judgment about newsworthiness and accuracy verification at point-of-transmission is often valued and legally protected through editorial oversight.
Adoption barriersclaude-sonnet-52/5No licensing barrier, but physical presence in remote locations and equipment operation create practical barriers to any automated substitution.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI integration for remote transmission logistics requires significant human oversight, fallback human handling when connectivity fails, and specialized infrastructure setup. The all-in cost (infrastructure, integration, human monitoring) remains comparable to or exceeds direct human transmission work in the field.
Cost vs. human wageclaude-sonnet-52/5AI offers no direct substitute for the physical act of transmitting via satellite phone or fax, so there's no meaningful AI cost comparison—human labor and equipment costs dominate.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can draft text and schedule automated transmissions in controlled environments, no deployed product reliably handles the full context of remote transmission—equipment selection, connectivity troubleshooting, adaptive protocol choice, and real-time decision-making in the field. Deployed systems exist for structured data transfer only.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product handles physical transmission equipment operation or field logistics from remote locations; this remains a human/physical task with basic digital tools.

Arrange interviews with people who can provide information about a story.

21

CI 11–30 · exposure 13 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Newsrooms have adopted AI for writing assistance and research, but interview arrangement remains highly personalized and relationship-dependent. Adoption of AI for source contact is slow because it conflicts with journalistic practice and requires human follow-up; few organizations deploy automation here.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for research and drafting but the interview-arranging function itself remains largely manual and relationship-driven, with slow uptake.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by generating lists of potential sources, summarizing their backgrounds, and drafting initial contact templates, which saves journalists research time. However, the critical work of building trust and negotiating access remains fundamentally human.
Augmentation potentialclaude-sonnet-53/5AI can help identify potential sources, draft outreach messages, and manage scheduling logistics, meaningfully assisting parts of the task even though the core interpersonal work remains human.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with some logistics (identifying potential sources, drafting interview request emails), but the task requires nuanced judgment about credibility, availability, and relationship-building that heavily depends on human outreach and persuasion. Current systems cannot reliably negotiate or close interviews end-to-end.
Task automatabilityclaude-sonnet-51/5Arranging interviews requires identifying appropriate sources, building rapport, negotiating access, and often persuading reluctant subjects—social and relational work AI cannot perform end-to-end today.'
Adoption barriersclaude-haiku-4-5-202510014/5Journalism operates under strong professional norms around direct relationship-building, trust, and accountability with sources. Editorial standards and source verification require human judgment; there is also implicit pressure to maintain personal relationships and reputation in journalism communities.
Adoption barriersclaude-sonnet-53/5No licensing barrier exists, but sources generally expect and prefer human contact for trust and journalistic credibility, creating social friction against automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5An AI system capable of autonomous source-finding and outreach would require significant human oversight, vetting, and relationship management, limiting cost savings. A journalist's time on outreach remains cheaper than building and maintaining bespoke AI systems for this interpersonal task.
Cost vs. human wageclaude-sonnet-52/5AI could cheaply draft outreach emails or research contact lists, but the actual negotiation and relationship-building still requires paid human time, limiting cost savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate contact lists and draft outreach messages, no deployed product reliably arranges interviews at scale. Humans still conduct the actual outreach, negotiation, and scheduling; AI tools exist only for research and drafting support, not for autonomous execution.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously identifies, contacts, and secures interview commitments with human sources; this remains a research-stage capability at best.

Investigate breaking news developments, such as disasters, crimes, or human-interest stories.

19

CI 7–30 · exposure 13 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While newsrooms increasingly use AI for story sorting and summarization, adoption of AI-driven investigation of breaking news is still limited to pilots and internal tools. Most breaking news reporting remains human-led, and few organizations have deployed autonomous or near-autonomous investigation systems.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for transcription, summarization, and social monitoring, but the core on-scene investigative task remains largely untouched by production AI systems.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants can significantly augment reporters by rapid background research, organizing leads, summarizing documents, and flagging relevant information, allowing journalists to focus on field reporting and source interviews. This is already common in newsrooms using AI writing and research aids.
Augmentation potentialclaude-sonnet-53/5AI can assist with monitoring social media/scanners for breaking news alerts, transcribing interviews, and organizing research, providing moderate productivity gains to human reporters.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with some elements (initial research, summarizing leads, scanning public records), but investigating breaking news requires on-scene reporting, source cultivation, verification under time pressure, and judgment about newsworthiness that humans must direct. Meaningful automation of the full investigation process to 50% time-saving remains unfeasible.
Task automatabilityclaude-sonnet-51/5Breaking news investigation requires physical presence, real-time human sourcing, on-scene observation, and building trust with witnesses/officials—none of which current AI can perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Editorial responsibility, journalistic ethics (source protection, attribution), legal liability for errors, and audience expectations for human accountability create strong organizational and professional barriers. News organizations face reputational risk if automation replaces human investigation without appropriate human oversight.
Adoption barriersclaude-sonnet-53/5No licensing requirement to be a journalist, but strong reliance on human judgment, credibility, source relationships, and legal/ethical accountability for accurate reporting creates real friction against substitution.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference is cheap, but integration with newsgathering workflows, fact-checking oversight, and the need for human reporters to execute the actual investigation means the all-in cost remains high relative to marginal human reporter time saved on preliminary research.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for on-the-ground investigation, so cost comparison favors the human by default since the AI alternative doesn't exist as a functional replacement.
Technical feasibility todayclaude-haiku-4-5-202510012/5Current AI tools (language models, search) support research and drafting but cannot autonomously conduct investigations, conduct interviews, verify facts through original reporting, or navigate the ethical and legal constraints of journalism. No deployed product reliably performs the full investigative task.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently investigates disasters, crimes, or human-interest stories on-site; AI tools only assist with monitoring feeds or drafting after human-gathered facts.

Conduct taped or filmed interviews or narratives.

13

CI 5–21 · exposure 5 · augmentation 50 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5News organizations have shown minimal adoption of AI-conducted interviews in production; the sector remains cautious about automation of interview-gathering because it threatens credibility and brand trust. Adoption is limited to narrow, low-stakes templated content.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for transcription, research, and drafting, but not for conducting the actual interview itself, which remains a manual, in-person task.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist journalists by suggesting follow-up questions, transcribing and summarizing interviews, or auto-generating interview outlines, raising productivity in preparation and post-production. However, the core task of conducting the interview itself remains human-centric.
Augmentation potentialclaude-sonnet-53/5AI assists with interview prep, question generation, transcription, and post-production editing, meaningfully aiding the journalist without performing the interview itself.
Task automatabilityclaude-haiku-4-5-202510011/5Conducting interviews or narratives requires real-time interpersonal judgment, adaptive questioning, and rapport-building that current AI cannot replicate. AI lacks the embodied presence, emotional intelligence, and ability to follow spontaneous conversational threads needed for journalistic interviews.
Task automatabilityclaude-sonnet-51/5Conducting live taped or filmed interviews requires real-time human rapport-building, adaptive follow-up questioning, and physical presence with camera/audio crews that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: credibility and trust require human bylines, editorial liability for accuracy rests with human journalists, and audience expectations strongly favor human-conducted interviews. Regulatory and professional norms (journalism ethics codes) reinforce human involvement.
Adoption barriersclaude-sonnet-53/5No formal licensing requirement, but human presence, trust-building, and subject willingness to be interviewed by an AI create strong practical and reputational barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of setting up, monitoring, and editing AI-generated interview footage, plus human oversight to ensure accuracy and journalistic integrity, remains comparable to or exceeds the cost of hiring a reporter to conduct the interview directly.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate scripted video or conduct basic Q&A, no deployed product reliably conducts authentic, investigative, or narrative interviews at professional journalism standards. Existing tools produce stiff, predictable interactions that lack the nuance required for publication.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously conducts filmed interviews with human subjects; AI voice agents exist for narrow scripted calls but not journalistic on-camera interviewing.

Present live or recorded commentary via broadcast media.

13

CI 9–16 · exposure 5 · augmentation 38 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5News organizations remain deeply conservative on presenter automation; pilot projects have faced public backlash. Actual production displacement of broadcast journalists by synthetic presenters is negligible today, even in lower-tier outlets.
Sector adoption velocityclaude-sonnet-52/5Broadcast media has been slow to adopt AI presenters; use remains experimental and limited to a handful of markets, with strong audience and industry resistance to fully AI-driven commentary.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with script generation, teleprompter text drafting, and real-time fact-checking overlays, but these are peripheral to the core act of presenting commentary. The augmentation is limited and does not materially transform broadcaster productivity.
Augmentation potentialclaude-sonnet-53/5AI assists with teleprompter scripting, real-time translation, data visualization, and research support behind the scenes, improving efficiency while humans still deliver the live commentary.
Task automatabilityclaude-haiku-4-5-202510011/5Presenting live or recorded commentary via broadcast media requires authentic human voice, credibility, emotional resonance, and real-time interaction with audience or co-hosts. Current AI cannot convincingly replace the human presence, judgment, and rapport central to broadcast journalism.
Task automatabilityclaude-sonnet-51/5Live/recorded on-camera or on-air commentary requires human presence, voice, likeness, and perceived credibility; AI avatars/voices exist but do not substitute for the actual presenting task at scale in real newsrooms today.》
Adoption barriersclaude-haiku-4-5-202510014/5Broadcast journalists operate under FCC regulations, journalistic ethics codes, and audience trust expectations that legally and culturally require human attribution and accountability. News organizations face reputational and liability risk from automated or synthetic presentation without explicit disclosure.
Adoption barriersclaude-sonnet-54/5Broadcast journalism carries strong norms around credibility, trust, on-air identity, and in some markets regulatory/labor union rules protecting human presenters, creating significant adoption friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI text-to-speech and video synthesis are cheap, but integration, content oversight, deepfake detection safeguards, and legal review add significant cost. The marginal cost advantage over hiring human presenters is modest when compliance and brand risk are factored in.
Cost vs. human wageclaude-sonnet-52/5Synthetic avatar/voice generation is cheap per minute of content, but production, legal, and quality-control overhead for credible live commentary keeps overall costs comparable to or not dramatically below human talent costs.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate synthetic speech and deepfakes exist, no deployed broadcast product reliably performs this task as a substitute for human presenters. Existing systems lack the authenticity, regulatory acceptance, and audience trust needed for production journalism.
Technical feasibility todayclaude-sonnet-51/5Some experimental AI news anchors exist (e.g., in China, Kuwait) but these are novelty/limited-scale deployments, not reliable mainstream production replacements for human broadcast commentary.

Establish and maintain relationships with individuals who are credible sources of information.

11

CI 5–16 · exposure 8 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Newsrooms have shown minimal adoption of AI for source relationship management because the task is fundamentally relational and reputational; adoption remains limited to AI-assisted discovery, not autonomous relationship maintenance.
Sector adoption velocityclaude-sonnet-52/5Newsrooms are adopting AI for drafting and research support, but relationship-building with sources remains untouched by adoption trends.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by identifying sources, analyzing their credibility history, suggesting contact strategies, and managing relationship databases, moderately boosting a journalist's ability to maintain a diverse source network while the human drives trust-building.
Augmentation potentialclaude-sonnet-53/5AI can help journalists track contacts, summarize past interactions, or prep questions, offering moderate support to the relationship management process without doing the relational work itself.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can identify potential sources and draft outreach messages, building genuine trust relationships requires sustained human interaction, judgment about credibility nuance, and personality fit that current systems cannot replicate end-to-end. Automation covers only preparatory parts, not the core relationship-building.
Task automatabilityclaude-sonnet-51/5Building trust-based, ongoing human relationships requires social presence, reputation, and reciprocity that AI cannot substitute for; no current system can originate or sustain source relationships.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: journalists' professional ethics codes require personal accountability for source relationships, legal liability for source protection falls on the human, and sources themselves demand direct human contact and trust—effectively requiring a licensed human be responsible.
Adoption barriersclaude-sonnet-54/5Source trust, confidentiality, and reputational/legal risk (e.g., protecting anonymous sources) create strong practical and ethical barriers to any automated substitute.
Cost vs. human wageclaude-haiku-4-5-202510011/5Replacing human journalists in source relationship management would still require human oversight and relationship repair; the cost of errors (burned sources, lost credibility) far exceeds typical AI inference costs, making full automation economically unfavorable.
Cost vs. human wageclaude-sonnet-51/5There is no AI alternative performing this task, so cost comparison favors the human by default since AI cannot substitute at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably establishes and maintains journalist-source relationships autonomously. AI can assist with source discovery and initial contact, but no production system performs the trust-building and maintenance relationship work that defines this task.
Technical feasibility todayclaude-sonnet-51/5No deployed product establishes or maintains journalistic source relationships; this remains entirely a human interpersonal function.

Discuss issues with editors to establish priorities or positions.

10

CI 5–15 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Newsrooms are conservative adopters of AI for core editorial decision-making; this task sits at the core of journalistic autonomy and editorial independence, with no evidence of meaningful automation adoption in practice.
Sector adoption velocityclaude-sonnet-53/5Newsrooms are adopting AI tools for drafting, research, and transcription at a moderate pace, but the core editorial negotiation process itself sees minimal AI penetration.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist by preparing briefing materials or summarizing data ahead of discussions, but the substantive conversation itself—establishing positions and negotiating priorities—depends on human editorial expertise and authority that AI cannot materially augment.
Augmentation potentialclaude-sonnet-53/5AI can help prepare briefing materials, summarize competing coverage, or surface data-driven angles to inform the discussion, but it does not substitute for the deliberation itself.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires dynamic two-way dialogue with human editors to negotiate priorities and positions based on nuanced editorial judgment. Current AI systems cannot reliably conduct substantive strategic discussions with human stakeholders or independently establish newsroom priorities.
Task automatabilityclaude-sonnet-51/5This is an interactive, judgment-driven human dialogue about editorial priorities and positions that requires organizational context, relationship dynamics, and real-time negotiation AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Editorial decisions involve significant liability, brand reputation risk, and newsroom authority structures where human editors bear legal responsibility for published content. Organizational norms and professional journalism standards also expect human editorial judgment and accountability in priority-setting.
Adoption barriersclaude-sonnet-53/5No formal licensing barrier exists, but strong organizational and editorial-judgment norms, accountability for editorial decisions, and the inherently interpersonal nature of the task create real friction against automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5An AI system capable of meaningful editorial discussions would require extensive customization, oversight, and integration with newsroom workflows—making it more expensive than having journalists conduct these conversations directly with editors.
Cost vs. human wageclaude-sonnet-51/5There is no AI product performing this function, so no meaningful cost comparison exists; the human interaction is irreplaceable at present.
Technical feasibility todayclaude-haiku-4-5-202510011/5No production system today reliably performs editorial priority-setting discussions autonomously. While AI can summarize talking points or draft agendas, the actual negotiation and decision-making with editors requires human editorial authority and judgment.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for an editor-reporter discussion establishing editorial stance or priorities; this remains a human collaborative process.

Coordinate and serve as an anchor on news broadcast programs.

8

CI 5–11 · exposure 0 · augmentation 38 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI anchors in production is minimal; most major news organizations maintain human anchors despite cost pressures, reflecting regulatory requirements, audience trust concerns, and liability risks that slow any shift toward automation.
Sector adoption velocityclaude-sonnet-52/5Media/broadcasting has seen slow, cautious AI adoption for anchoring specifically, with only a few high-profile experiments and no widespread rollout of AI-only broadcast anchors.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with script generation, fact-checking, or segment timing cues, but these are peripheral to the core task of on-air anchoring, which remains fundamentally human-driven; augmentation potential is limited to preparation phases, not performance.
Augmentation potentialclaude-sonnet-53/5AI can help anchors with teleprompter scripting, research summarization, translation, and rehearsal, meaningfully aiding production despite not replacing the live hosting role.
Task automatabilityclaude-haiku-4-5-202510011/5Anchoring a live news broadcast requires real-time presence, vocal performance, editorial judgment on-air, and audience engagement that current AI cannot replicate end-to-end. While AI can draft scripts or suggest segment arrangements, the actual performance of being an on-air anchor remains a fully human task.
Task automatabilityclaude-sonnet-51/5Live anchoring requires real-time on-camera presence, spontaneous ad-libbing, interviewing, and handling breaking news dynamically, none of which current AI can perform end-to-end with equal quality.
Adoption barriersclaude-haiku-4-5-202510014/5Broadcast anchoring faces significant barriers including FCC regulatory responsibilities, editorial liability for on-air statements, union representation (AFTRA), and strong audience expectations for human authenticity and trusted human judgment during news delivery.
Adoption barriersclaude-sonnet-54/5Broadcast journalism carries strong norms around human accountability, credibility, and audience trust in a named individual, plus regulatory/editorial standards that make full AI substitution highly resisted even if not strictly licensed.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying AI systems (video synthesis, speech generation, oversight, content moderation, corrections) to match a human anchor's output quality and credibility likely exceeds the salary of an entry-level anchor, and far below that of experienced anchors.
Cost vs. human wageclaude-sonnet-52/5While synthetic avatar generation is cheap per hour of content, the production, oversight, scripting, and legal/trust infrastructure needed to make it viable keeps effective costs comparable to or above human anchors for credible news.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs live news anchoring in production. While AI can generate video deepfakes or synthesize speech, these lack the editorial authority, real-time responsiveness, and viewer trust that define professional broadcast anchoring today.
Technical feasibility todayclaude-sonnet-51/5AI news avatars exist in limited experimental deployments (e.g., some Asian state media), but no mainstream broadcaster relies on AI as a primary live anchor for coordinating and hosting programs.

Related occupations — Arts, Design, Entertainment, Sports & Media

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.