Google organic-search referrals to more than 2,500 news sites fell 33% worldwide, and 38% in the United States, between November 2024 and November 2025 (Reuters Institute, Digital News Report 2026). That collapse defines AI’s arrival in journalism far more than any newsroom pilot: audiences are getting answers without clicking through. Only 10% of people use AI chatbots for news weekly, and just 20% trust the answers those chatbots give, yet 82% of journalists now use at least one AI tool in their own work (Muck Rack, 2026 State of Journalism Report). Meanwhile NewsGuard counts 3,749 unreliable AI-generated news sites polluting the same information space. This analysis consolidates data from the Reuters Institute, Pew Research Center, WAN-IFRA, Muck Rack, NewsGuard, and 10 other primary sources across audience behavior, trust, newsroom adoption, misinformation, and publisher economics.
TL;DR
- 10% of people globally use AI chatbots for news weekly in 2026, up from 7% in 2025 (Reuters Institute, Digital News Report 2026).
- The share of journalists using no AI tools at all fell from 33% to 21% in a year (Cision, 2026 State of the Media Report).
- 66% of US adults are highly concerned about inaccurate AI-generated information (Pew Research Center, 2025).
- 50% of US adults expect AI to have a negative effect on news; only 10% expect a positive one (Pew Research Center, 2025).
- 82% of journalists use at least one AI tool, up from 77% a year earlier (Muck Rack, 2026 State of Journalism Report).
- 44% of publishers call their AI results promising, but only 13% call them transformational (Reuters Institute, Trends and Predictions 2026).
- NewsGuard tracked 3,749 unreliable AI-generated news sites across 16 languages by June 2026, up from 2,089 in October 2025 (NewsGuard, AI Tracking Center).
- Google users click a link 8% of the time when an AI summary appears, versus 15% without one (Pew Research Center, 2025).
- Only 20% of publishers expect AI licensing to become a substantial revenue source (Reuters Institute, Trends and Predictions 2026).
- The AI in media market is projected to grow from $18.47B in 2026 to $60.35B by 2030 (The Business Research Company).
1. How Audiences Use AI for News
Adoption of AI chatbots for news is real but shallow. Weekly use climbed to 10% globally in 2026 from 7% a year earlier, yet only 1% of people name AI as their main source of news (Reuters Institute, Digital News Report 2026). The pattern is generational and geographic: under-35s and a handful of early-adopter markets pull the average up, while the US, UK, France, and Germany showed no year-on-year increase at all. For the broader search-behavior shift behind these numbers, see our AI search statistics for 2026.
| Metric | Value | Source |
|---|---|---|
| Use AI chatbots for news weekly (2026) | 10% (from 7% in 2025) | Reuters Institute, Digital News Report 2026 |
| Under-35s using AI chatbots for news | 16% | Reuters Institute, Digital News Report 2026 |
| Say AI chatbots are their main news source | 1% | Reuters Institute, Digital News Report 2026 |
| Weekly-use range across markets | 14% (South Korea) to 4% (UK) | Reuters Institute, Digital News Report 2026 |
Context: the Digital News Report surveys 48 markets with roughly 2,000 respondents each, fielded in early 2026.
2. Trust and Comfort With AI-Generated News
Trust is the ceiling on AI news adoption, and it is low. Only 20% of people globally trust what AI chatbots tell them, compared with 37% who trust the news in general (Reuters Institute, Digital News Report 2026). US attitudes are harsher still: in a Pew Research Center survey of 5,410 adults, half expected AI to worsen the news, two-in-three were highly concerned about inaccurate AI-generated information, and more people said AI writes worse stories than journalists than said it writes better ones.
| Metric | Value | Source |
|---|---|---|
| Trust answers from AI chatbots | 20% | Reuters Institute, Digital News Report 2026 |
| Trust in the news overall | 37% | Reuters Institute, Digital News Report 2026 |
| US adults expecting a negative AI effect on news | 50% | Pew Research Center, 2025 |
| US adults expecting a positive effect | 10% | Pew Research Center, 2025 |
| Say AI writes worse news stories than journalists | 41% (vs 19% better) | Pew Research Center, 2025 |
| Extremely or very concerned about inaccurate AI news | 66% | Pew Research Center, 2025 |
| Expect AI to reduce journalist jobs over 20 years | 59% | Pew Research Center, 2025 |
Flag: the Pew data was fielded August 2024 and published in 2025 (most recent US measure of AI-and-news attitudes at this granularity).
3. Inside the Newsroom: Adoption and Results
Publishers have moved from experimentation to workflow, but value is uneven. Only 13% of newsroom leaders call their AI results transformational, while 42% describe the impact as limited so far (Reuters Institute, Trends and Predictions 2026, 280 leaders across 51 countries). The clearest wins are operational, not commercial: WAN-IFRA’s publisher survey found efficiency gains far outrun direct revenue. Some outlets now auto-generate audio versions of articles and translations, but back-end automation, not audience-facing generation, is where adoption is deepest.
| Metric | Value | Source |
|---|---|---|
| Publishers rating AI results promising | 44% | Reuters Institute, Trends and Predictions 2026 |
| Rating AI results transformational | 13% | Reuters Institute, Trends and Predictions 2026 |
| Consider back-end automation important | 97% | Reuters Institute, Trends and Predictions 2026 |
| Newsrooms rating AI adoption advanced | 45.5% | WAN-IFRA, World Press Trends Outlook 2025-2026 |
| Publishers reporting efficiency gains from AI | 75% | WAN-IFRA, 6th AI Report 2025 |
| Reporting better content production | 64% | WAN-IFRA, 6th AI Report 2025 |
| Reporting direct revenue gains from AI | 9% | WAN-IFRA, 6th AI Report 2025 |
| Publishers reporting no jobs cut due to AI | 67% | Reuters Institute, Trends and Predictions 2026 |
Context: WAN-IFRA’s value survey covered 100+ media leaders in Q2 2025; the adoption breakdown draws on 170+ executives across 66 countries.
4. Journalists’ Own AI Usage and the Labor Squeeze
Adoption among individual journalists is now the majority position, driven partly by shrinking newsrooms. 82% of journalists use at least one AI tool, up from 77% a year earlier, and the share using no AI at all fell from 33% to 21% (Muck Rack, 2026 State of Journalism Report; Cision, 2026 State of the Media Report). But usage and unease are rising together: more journalists flag AI as a top challenge than did last year. The same tooling pressure shows up on the PR side, which we cover in our AI in marketing statistics for 2026.
| Metric | Value | Source |
|---|---|---|
| Journalists using at least one AI tool | 82% (from 77% in 2025) | Muck Rack, 2026 State of Journalism |
| Journalists using ChatGPT | 47% | Muck Rack, 2026 State of Journalism |
| Journalists using transcription tools | 40% | Muck Rack, 2026 State of Journalism |
| Cite unchecked AI as a top concern | 26% (from 18%) | Muck Rack, 2026 State of Journalism |
| Journalists not using AI at all | 21% (from 33% in 2025) | Cision, 2026 State of the Media |
| Rank AI among their biggest challenges | 43% (from 30%) | Cision, 2026 State of the Media |
| Use AI to transcribe interviews and audio | 23% | Cision, 2026 State of the Media |
| Media job cuts, year over year | +18% | Cision, 2026 State of the Media |
Context: Muck Rack analyzed 897 responses (Jan-Mar 2026); Cision surveyed 1,800+ journalists across 19 markets.
5. AI Content Farms and Misinformation
The supply side of AI news is expanding faster than the demand side. NewsGuard tracked 3,749 unreliable AI-generated news sites across 16 languages by June 2026, up from 2,089 in October 2025, an increase of more than 900 sites in roughly five months (NewsGuard, AI Tracking Center). These operations publish with minimal human oversight and no disclosure, and quality audits find chatbots themselves repeating false claims on news topics more than a third of the time. For the detection side of this arms race, see our deepfake detection statistics for 2026.
| Metric | Value | Source |
|---|---|---|
| Unreliable AI-generated news sites tracked | 3,749 | NewsGuard, AI Tracking Center |
| Languages covered | 16 | NewsGuard, AI Tracking Center |
| Sites tracked, October 2025 | 2,089 | NewsGuard, AI Tracking Center |
| Sites tracked, March 2026 | 3,006 | NewsGuard, AI Tracking Center |
| New sites added in roughly five months | 900+ | NewsGuard, AI Tracking Center |
| AI chatbots repeating false claims on news topics | more than 33% of prompts | NewsGuard, AI Tracking Center |
| Journalists citing disinformation as a top threat | 32% | Muck Rack, 2026 State of Journalism |
| Journalists saying misinformation complicated their work | over 50% | Muck Rack, 2026 State of Journalism |
Outlier: NewsGuard’s site count roughly doubled in eight months, marking a shift from viral one-off fakes toward industrialized content production.
6. Publisher Economics: Licensing and the Search Traffic Collapse
The core business trade of 2026 is licensing revenue against referral loss, and the loss is winning. Google organic-search referrals to 2,500+ news sites fell 33% globally and 38% in the US year over year, and publishers expect a further 43% decline over three years (Reuters Institute, Trends and Predictions 2026). Pew’s browsing-panel data explains the mechanism: when an AI summary appears, users click a link far less often. Licensing deals help a few large publishers, but most see them as marginal.
| Metric | Value | Source |
|---|---|---|
| Google organic-search referrals to news sites, YoY | -33% global / -38% US | Reuters Institute, Digital News Report 2026 |
| Publisher-expected search referral decline (3 yrs) | -43% | Reuters Institute, Trends and Predictions 2026 |
| Link clicks with vs without an AI Overview | 8% vs 15% | Pew Research Center, 2025 |
| Searches ending with no click when AI summary shown | 26% (vs 16%) | Pew Research Center, 2025 |
| News Corp-OpenAI licensing deal value | $250M+ over 5 years | Quartz / Digiday reporting, 2026 |
| Reddit content licensing (reported) | ~$60M per year | Quartz / Digiday reporting, 2026 |
| Publishers listing AI licensing as a revenue priority | 37% | Reuters Institute, Trends and Predictions 2026 |
| Publishers expecting licensing to be substantial revenue | 20% | Reuters Institute, Trends and Predictions 2026 |
Context: OpenAI alone has signed content deals with 20+ news publishers covering 160+ outlets in over 20 languages; disclosed deal values range from roughly $5M to $250M (Quartz).
7. Market Size and Outlook
Sizing this market depends heavily on scope, so cross-referencing matters. Analysts put the broad AI in media market at $18.47 billion in 2026, on track for $60.35 billion by 2030 at a 34.5% CAGR (The Business Research Company). The narrower automated-journalism segment is far smaller and estimates diverge sharply. On the value side, journalism remains disproportionately important to AI systems: Muck Rack finds journalistic sources account for 27% of the links generative AI cites, rising to nearly half on recency-driven queries.
| Metric | Value | Source |
|---|---|---|
| AI in media market, 2026 | $18.47B | The Business Research Company, 2026 |
| AI in media market, 2030 (projected) | $60.35B | The Business Research Company, 2026 |
| AI in media market CAGR, 2026-2030 | 34.5% | The Business Research Company, 2026 |
| Automated journalism market, 2025 | $1.61B | Verified Market Reports, 2026 |
| Automated journalism market, 2034 (projected) | $5.14B (14.1% CAGR) | Verified Market Reports, 2026 |
| AI training dataset market, 2026 | $4.44B | Fortune Business Insights, 2026 |
| Journalism’s share of links cited by generative AI | 27% | Muck Rack, Generative Pulse 2026 |
| Journalism’s share for recency-implied queries | 49% | Muck Rack, Generative Pulse 2026 |
Divergence: a separate estimate sizes automated journalism at just $180M in 2025 scaling to $1.11B by 2035 (Market Research Intellect), a reminder that any single market number should be treated as directional.
Summary: AI In Journalism by the Numbers
| Metric | Value | Source |
|---|---|---|
| Use AI chatbots for news weekly (2026) | 10% (from 7%) | Reuters Institute, Digital News Report 2026 |
| Trust answers from AI chatbots | 20% | Reuters Institute, Digital News Report 2026 |
| Highly concerned about inaccurate AI news info | 66% | Pew Research Center, 2025 |
| Newsrooms rating AI adoption advanced | 45.5% | WAN-IFRA, World Press Trends 2025-2026 |
| US adults expecting AI to harm the news | 50% | Pew Research Center, 2025 |
| Expect AI to reduce journalist jobs (20 yrs) | 59% | Pew Research Center, 2025 |
| Journalists using at least one AI tool | 82% | Muck Rack, 2026 State of Journalism |
| Journalists using no AI at all | 21% (from 33%) | Cision, 2026 State of the Media |
| Publishers calling AI results transformational | 13% | Reuters Institute, Trends and Predictions 2026 |
| Publishers reporting efficiency gains from AI | 75% | WAN-IFRA, 6th AI Report 2025 |
| Publishers reporting direct revenue gains | 9% | WAN-IFRA, 6th AI Report 2025 |
| Unreliable AI-generated news sites tracked | 3,749 | NewsGuard, AI Tracking Center |
| Google organic-search referrals to news, YoY | -33% global / -38% US | Reuters Institute, Digital News Report 2026 |
| Link clicks with vs without an AI Overview | 8% vs 15% | Pew Research Center, 2025 |
| Publishers expecting licensing to be substantial revenue | 20% | Reuters Institute, Trends and Predictions 2026 |
| News Corp-OpenAI deal value | $250M+ over 5 years | Quartz / Digiday reporting, 2026 |
| AI in media market, 2026 | $18.47B | The Business Research Company |
| AI in media market, 2030 (projected) | $60.35B | The Business Research Company |
| Journalism’s share of links cited by generative AI | 27% | Muck Rack, Generative Pulse 2026 |
Methodology and Sources
Data was gathered by aggregating figures directly from primary reports, surveys, market-research firms, and named trackers published in 2025 and 2026, with market-size and traffic figures cross-referenced across two or more sources where possible.
Sources cited:
- Reuters Institute for the Study of Journalism, Digital News Report 2026 (48 markets, ~2,000 respondents each)
- Reuters Institute, Journalism, Media and Technology Trends and Predictions 2026 (280 leaders, 51 countries)
- Pew Research Center, Americans Largely Foresee AI Having Negative Effects on News (2025) (5,410 US adults)
- Pew Research Center, Google Users Click Less When AI Summaries Appear (2025) (900 US adults, browsing panel)
- WAN-IFRA, 6th AI Report: The AI Value Equation (2025) and World Press Trends Outlook 2025-2026
- WAN-IFRA, AI at Work: How Newsrooms Are Redefining Production (2026)
- Muck Rack, 2026 State of Journalism Report (897 responses) and Generative Pulse (2026)
- Cision, 2026 State of the Media Report (1,800+ journalists, 19 markets)
- NewsGuard, AI Tracking Center / Unreliable AI-Generated News Sites
- The Business Research Company, AI in Media Market (2026)
- Verified Market Reports, Automated Journalism Market (2026); Market Research Intellect (divergent estimate)
- Fortune Business Insights, AI Training Dataset Market (2026)
- Quartz and Digiday reporting on AI publisher licensing deal values (2026)
Data watch: The Reuters Institute publishes its Digital News Report every June and its Trends and Predictions each January, so both will have fresh 2027 editions within months. Muck Rack (State of Journalism) and Cision (State of the Media) run annually in Q1, WAN-IFRA updates its AI reports through the year, and NewsGuard revises its AI Tracking Center continuously. Pew Research periodically re-fields its news-and-AI attitudes questions; the current US figures were collected in 2024.
Last updated: July 10, 2026.
We review and update this page quarterly as new data is published.