Reddit and publishers weigh blocking Google over AI traffic drops

For decades, the deal between Google and the web was simple: publishers let Google crawl their pages, and Google sent traffic back.

AI summaries have shattered that arrangement. Instead of passing readers through to original sources, search engines now synthesize the answers right on the results page. The content gets used, but the clicks never arrive.

According to The Wall Street Journal, a growing list of heavyweights—including Reddit, Politico, USA Today, Reuters, and The Economist—are now weighing whether to block Google's web crawler entirely.

The traffic cliff hits publishers hard

Publishers are considering extreme measures because the referral numbers are plummeting.

Data from Semrush highlights just how steep the drop has been over the past year. USA Today saw its Google search traffic slide by 18% year-over-year, while Politico experienced a 20% decline.

For others, the bleed is even worse. Both CNN and Business Insider recorded search traffic drops of 31% over the same period.

When Google answers queries directly using publishers' reporting, users have little reason to click outside links. That directly erodes the pageviews that sustain ad-supported journalism.

Reddit's $60 million dilemma

The most notable name in the revolt is Reddit, which previously embraced Google's AI push.

Back in 2024, Reddit signed a high-profile licensing deal worth $60 million annually, giving Google access to its vast user forums to train AI models like Gemini.

With that agreement now approaching renewal, Reddit executives are openly questioning the upside. Sources told The Wall Street Journal that the company has discussed cutting off Google's access to training data, wondering what value remains when AI answers steadily cannibalize search referrals.

As The Verge noted, this talk could easily be a negotiation bluff. Threatening to walk away right before contract talks is a classic playbook to extract higher licensing fees.

Why it matters

We think this moment marks the end of the uneasy truce between Big Tech and digital publishers.

Whether Reddit is posturing or genuinely considering a breakup, the underlying tension is identical everywhere: tech platforms are using open web data to build systems that eliminate the need for the open web. The so-called "Google Zero" future, where search engines keep all the traffic for themselves, is arriving faster than anticipated.

For everyday internet users, this standoff carries real downsides. If trusted newsrooms and community forums lock their doors against scrapers or retreat behind hard paywalls, the web becomes far more fragmented. Search quality degrades, and AI models are left training on synthetic slop rather than human reporting.

Publishers face an unenviable gamble. Blocking Google risks an immediate and painful traffic drought, but continuing to feed the crawler looks like slow-motion surrender. The terms hammered out in these upcoming renewals will establish the price tag for the entire AI data economy.