Publishers Weigh Pulling Their Content From Google’s AI

The bargain that built the open web — publishers let Google crawl their pages, and Google sent readers back — is starting to come apart, and some of the largest names in online media are now weighing whether to keep supplying the company at all.

Reddit has gone furthest, debating whether to withhold its content from Google’s AI altogether, according to the Wall Street Journal, which reported that USA Today (TDAY ) (TDAY ), Politico, the Economist, People Inc. and Reuters are also evaluating how, or whether, to continue working with the company. The frustration has built as Google expands AI-generated answers that resolve a query on the results page and leave publishers with the lost visit but not the click.

The tension is sharpest at Reddit because Reddit had made its peace. In 2024 the platform signed a licensing deal, reported at about $60 million a year, that gave Google structured access to its data feed to train and ground AI models on the site’s discussions. That agreement is ending soon, and the two sides are in renewal talks. With AI answers cutting clicks to outside sites, Reddit executives are reassessing what they gain from continuing to feed Google, even as the platform has become one of the most-cited sources in AI answers.

Why the leverage is shifting

The traffic data explains the anger. Google introduced AI Overviews, the summaries that sit above search results, to all US users in May 2024, and the effect on referrals became measurable within a year. A Pew Research Center study found that when an AI Overview appears, users click through to a traditional search result only 8% of the time, against 15% when no summary is shown. Chartbeat data cited by the Reuters Institute put the global drop in Google search referrals to publishers at roughly a third over the year to November 2025. That erosion has been central to publishers’ fear for the open web.

For years, publishers had little practical recourse. Opting out of Google’s AI meant opting out of Google Search: the same controls that blocked AI crawlers also stripped a site of the snippets that keep it visible in ordinary results. Vanishing from the index was not survivable for most outlets, so withholding content was no real choice. That is why a credible threat to walk away has belonged only to the handful of platforms, Reddit among them, whose data is genuinely hard to replace.

The regulator that changed the math

What changed the calculation was a regulatory intervention. On June 3, 2026, the UK’s Competition and Markets Authority imposed a new rule on Google that, for the first time, lets publishers stop their content from powering AI features such as AI Overviews while still appearing normally in search. Google is barred from downranking or otherwise penalizing sites that opt out, and the carve-out reaches the fine-tuning of models as well as the live answers. The regulator also requires Google to attribute publisher content with clear links inside AI-generated results.

The move came after the watchdog designated Google a company with “strategic market status” — its term for entrenched market power that unlocks bespoke rules — in general search, where Google handles more than 90% of UK queries. The point, it said, is bargaining power. “It is crucial that content publishers, including news organisations, have appropriate bargaining power over how their content is used,” said CMA chief executive Sarah Cardell. Google has nine months to comply and must file regular reports showing how.

That rule is the machinery now sitting behind the publishers’ revolt. A believable threat to cut Google off depends on being able to leave AI without leaving search, and in Britain that ability now exists in law.

The US picture, and what to watch

Publishers in the United States, home to most of the companies named this week, have no equivalent opt-out mandate, so their pressure runs through the courts. Penske Media, which owns Rolling Stone, Billboard and Variety, sued Google in 2025, arguing that AI Overviews cannibalize the traffic publishers were promised in return for being indexed; Google has moved to dismiss the case, and that motion is pending before the court. The New York Times (NYT ) is pressing a separate copyright case against OpenAI and Microsoft (MSFT ) (MSFT ), and Google itself was ruled an illegal search monopoly in a federal case it is now appealing.

Brussels is pushing from another direction, having issued binding instructions under its digital-competition rules to loosen Google’s grip on search and Android. The common thread across all three capitals is that policymakers are trying to manufacture the leverage that the old crawl-for-traffic bargain no longer supplies on its own.

Whether the publishers follow through is the open question. Reddit has signaled it wants more than a flat fee, pushing for pricing that rises as its content becomes more central to AI answers, and its renewal talks will show how much a single high-value supplier can extract. The CMA’s compliance clock, meanwhile, will reveal whether opt-out rights turn into paid licensing or merely into lost reach. For the companies that spent two decades optimizing for Google, the telling fact is that they are, for the first time, seriously pricing out life without it.