News publishers are seeing their search and social lifelines fray at the same time—and AI is a big part of the story. New analysis from the Reuters Institute for the Study of Journalism, drawing on traffic data from Chartbeat across 2,756 outlets worldwide, finds Google search referrals to news sites fell 33% globally and 38% in the US over the past year. The drop coincides with the spread of AI-generated answers in search, which increasingly satisfy users without a click.
Search Referrals Sink As AI Answers Rise
Google’s AI Overviews place multi-paragraph, machine-written summaries at the top of many results pages. The format pushes links below the fold or behind extra taps, which can depress click-throughs. Chartbeat’s global view shows a consistent slide after AI summaries expanded, particularly outside the US. Inside the US, the trendline is choppier, with some months of recovery—underscoring that not every query, nor every publisher, is affected equally.
Notably, hard news topics appear less likely to trigger AI Overviews compared with lifestyle or utility queries like weather, TV listings, or horoscopes. That uneven impact matters: publishers heavy on “utility news” may be losing more search visibility than those focused on breaking or investigative coverage.
Google disputes the scale of the damage. The company says its internal metrics do not reflect the dramatic declines, questions the site sample, and notes that some figures exclude Google News click-throughs. It also points to recent changes meant to surface links more visibly within AI results and a “Preferred Sources” option in Google News that lets users prioritize outlets they trust. Even so, the topline numbers mark a sharp turn from the roughly 10% search decline reported by industry group Digital Content Next last year.
Discover Overtakes Search Then Stumbles
There’s more turbulence on mobile. Chartbeat’s data shows Google Discover—the personalized feed that appears on Android and within the Google app—now drives more publisher traffic than classic search, accounting for 13% of referrals worldwide versus 7.3% for search. But Discover is wobbling too: referrals fell 21% globally and 29% in the US year over year.
That matters because Discover is both powerful and opaque. A single card can deliver surges of readers, but placement depends on shifting signals: user interests, freshness, authority, and increasingly, how AI systems interpret engagement quality. For publishers that reoriented their workflows toward mobile feeds, volatility in Discover can feel like the floor moving underfoot.
Social Platforms Are No Safety Net for News Publishers
The social side is hardly a refuge. Since mid-2023, Chartbeat tracks Facebook referrals down 43% globally and 35% in the US, while X (formerly Twitter) has fallen 46% worldwide and 45% in the US. Some publishers did see a rebound over just the past year—Facebook referrals up 23% and X up 29% in the US—but the longer arc remains negative.
Policy and product shifts are biting. Facebook has steadily de-emphasized news in feeds and rolled back publisher programs. X has experimented with hiding headlines in link previews and has signaled that posts with external links may be downranked, reducing off-platform traffic. In some markets, regulatory frictions have also constrained news distribution on social platforms, further fragmenting audience flows.
Publishers Pivot To Video Strategies And AI Deals
Faced with shrinking referrals, news leaders are adjusting their bets. In a Reuters Institute survey of digital-media executives, plans to invest more in Facebook and Google search earned net scores of -23 and -25, respectively, with X at -52. By contrast, YouTube scored +74, AI platforms +61, and TikTok +56—signaling a tilt toward video ecosystems and channels where direct audience building still feels possible.
Licensing content to AI firms is emerging as a pragmatic hedge. Two in ten executives expect “substantial” revenue from such deals, while nearly half expect “minor” gains. The calculus is straightforward: if AI systems are reshaping discovery, publishers want compensation, attribution, and—ideally—preferential exposure within those AI surfaces.
Operationally, outlets are doubling down on owned channels (newsletters, apps, direct subscriptions) and on structured data strategies that help machines understand and credit their work. Clear bylines, precise topics, fast pages, and consistently updated evergreen guides matter more when ranking systems—and AI summarizers—reward authority and freshness signals.
What To Watch Next As AI Overviews Reshape Traffic
Three variables will define the next leg: how often AI Overviews appear on newsworthy queries, how prominently links are displayed inside those summaries, and whether Google brings more transparency to Google News and Discover referral accounting. If link visibility improves, some of the lost traffic could flow back. If not, expect further consolidation around platforms where audiences still click through and around direct relationships publishers own.
The broader takeaway is blunt but useful. Distribution has changed faster than newsroom playbooks. AI now sits between publishers and readers in search, while social platforms that once promised reach are throttling links. The winners will be those that diversify their acquisition mix, negotiate hard for AI-era economics, and make content that earns its click even when the answer box tries to do the job first.
