Google's Publisher Favor-Button Is a Defense Against AI Scrapers

The Gesture Looks Generous. It Isn’t.

Google on August 20 announced that publishers can now embed a “Preferred Sources” button directly on their websites, letting readers mark a site as a favorite that Google will highlight across Search, Discover, and Google News. On the surface, this reads like a lifecast thrown to an industry that has been watching its web traffic evaporate as AI-powered search features redirect queries away from landing pages and toward synthetic summaries.

But the timing and architecture of this feature tell a different story. Google is not simply rewarding publishers. It is building a walled garden around the very content that AI aggregators have been stripping bare—and positioning itself as the gatekeeper who decides which publishers deserve to survive.

The Real Threat Is Not AI Summaries. It Is AI Scrapers.

The public narrative around falling publisher traffic has focused on AI Overviews and Google’s AI Mode—features that answer queries directly without sending users to a website. That is only half the picture. The more dangerous competitor for publishers is the rising swarm of AI news aggregators that scrape, repackage, and redistribute original reporting without licensing deals or revenue splits.

Startups and platform companies have been quietly ingesting news content at scale. Some operate in gray zones between fair use and outright expropriation. They do not pay publishers. They do not attribute. They do not face consequences when their output degrades.

Google’s Preferred Sources program is, in effect, a signal-detection and prioritization system. By letting readers vote with a click, Google creates a ranked list of trusted sources that its algorithms can preferentially surface—and, just as importantly, can preferentially ignore. The 345,000 unique sources selected since the feature’s May rollout are not just engagement data. They are a crowd-sourced whitelist that Google can use to justify surfacing certain publishers and demoting others in both traditional search and AI-generated responses.

Who Wins and Who Loses

Publishers with brand recognition win. Large, established names—major newspapers, well-known outlets—are the ones readers will instinctively favorite. The click-through data Google cited (twice as likely when a preferred source is available) reinforces this feedback loop: famous outlets get more visibility, which brings more traffic, which justifies further investment in their presence. The advantage compounds.

Niche and regional publishers lose. A local newspaper, an independent newsletter, or a specialist trade outlet may have dedicated readers but will struggle to accumulate the volume of source selections needed to stay visible. The system rewards density of existing authority, not quality of coverage.

YouTube and Google News are the structural winners. This is the part English-language coverage usually misses. By tying the Preferred Sources signal to Google News, Discover, and increasingly to AI Mode—which surfaces content across multiple Google surfaces—Google is reinforcing the centrality of its own distribution channels. Publishers who embed the button are not just signaling preference; they are routing reader loyalty through Google’s infrastructure. YouTube, which already commands enormous news consumption time, benefits from the same algorithmic preference for verified sources when it surfaces video content.

AI scraper companies face growing exclusion risk. If Google’s systems begin treating publisher-favorite status as a ranking signal, unlicensed aggregators that pull content from non-preferred sources face a two-headed penalty: their outputs may rank lower in Search, and Google’s own AI summaries may cite fewer of them. The barrier to entry for scraping-based news products just rose, but not in a way that helps publishers—the benefit accrues to Google.

The Discover Feed Upgrade Is Bigger Than It Sounds

Alongside the button, Google announced that readers will soon be able to customize their Discover feed using natural language commands—telling the system what topics to show more or less of. This is a meaningful shift from the opaque, engagement-optimized curation that currently drives Discover. For the first time, readers are being given explicit control over what Google considers newsworthy.

Again, the second-order effect is the real story. When readers define their interests in plain language, Google’s models train on those explicit signals rather than inferred ones. The feedback loop becomes tighter. Publishers whose content aligns with stated preferences get amplified; those that do not get starved. The system does not reward breadth of coverage. It rewards alignment with reader-defined demand signals that Google controls.

Android users will also be able to customize audio daily briefings in the Google News app—a feature that extends personalization into the audio domain and further anchors the publisher-preference signal inside Google’s own apps rather than on the open web.

What This Means for Media Economics

The underlying math has not changed. AI-driven search continues to reduce click-through rates to publisher websites. No button will reverse that structural trend. What Google is offering is a partial mitigation: a way for favored publishers to recapture some traffic share that would otherwise dissipate across AI summaries and aggregator intermediaries.

But the long-term implication is subtler and more concerning. By formalizing a publisher-preference hierarchy, Google is effectively deciding which news organizations are legitimate and which are not—not through editorial judgment but through aggregate user clicks. This is governance by engagement metrics, and it favors outlets that already have scale, brand familiarity, and reader habits that translate into button presses.

Publishers should welcome the tool but resist the framing. The Preferred Sources button is a defensive feature, not an investment. It protects incumbents from being scraped and replaced, which is something the industry needs. But it does not create new revenue, it does not change the fundamental traffic decline, and it entrenches Google’s role as the arbiter of which publishers matter.

The smarter question is what happens next: will Google extend this signal into paid partnerships? Will preferred-source status become a monetizable asset that larger publishers can leverage? Or will it remain a free feature that quietly consolidates Google’s position as the unavoidable middleman between news and readers?

The answer will determine whether this feature helps publishers build independence or simply makes them more dependent on Google’s ecosystem.