Google Discover, the personalized content feed that surfaces articles and news for hundreds of millions of users, has a serious pollution problem. For months, users have been inundated with a deluge of low-quality, nonsensical, and often entirely fabricated “news” articles generated by AI. This isn’t just a minor annoyance; it’s a systemic failure that is eroding user trust and turning what was once a useful discovery tool into a digital wasteland of spam.
Now, after a wave of public complaints and investigative reports, Google has finally acknowledged the issue, stating they are “actively working on a fix.” But for the publishers who have been pushed out by this tide of AI-generated garbage and the users who are being misled, the question is not just when this fix will arrive, but whether it can possibly be enough to solve a problem that strikes at the very heart of Google’s ability to distinguish real content from synthetic spam.
Expert Analysis: “As SEOs and digital publishers, we’ve been watching this crisis unfold in slow motion. The AI spam in Discover isn’t a bug; it’s the inevitable result of a perfect storm: the commoditization of AI content generation and a gaping loophole in how Google treats expired domains. Spammers have turned this combination into a ‘cash machine,’ and right now, Google’s defenses are being completely overwhelmed. A simple ‘fix’ isn’t enough; this requires a fundamental rethinking of what content is allowed to be amplified.”
This BroadChannel deep dive explores the mechanics behind this spam epidemic, why Google’s current systems are failing, and whether the promised fix can truly save Google Discover from itself.
The Spam Playbook: How Scammers Hijacked Discover
The army of spammers flooding Google Discover is not using a single tactic but a sophisticated, multi-pronged strategy that exploits a specific vulnerability in Google’s algorithms.
Here’s the step-by-step playbook they are using to turn Discover into a firehose of low-quality content:
Step 1: The Expired Domain Heist
The entire operation is built on a foundation of expired domain abuse. Spammers use automated tools to find and purchase domain names that have recently expired but still retain some level of “authority” in Google’s eyes. These are often domains that belonged to legitimate businesses, non-profits, or even government agencies in the past. Because these domains have an established history and existing backlinks, Google’s algorithms are more likely to trust them than a brand-new domain.
Step 2: The AI Content Factory
Once they have acquired an authoritative domain, the spammers use AI content generation tools to populate it with hundreds or even thousands of articles at an incredible speed. This content is often poorly written, factually incorrect, and designed purely to target trending keywords and clickbait-style headlines. For example, a recent investigation by Press Gazette found AI-generated articles falsely claiming that the UK state pension age had been lowered or that over-60s could get free TV licenses—both entirely untrue.
Step 3: Infiltrating Discover
Because the content is hosted on a domain that Google’s systems already trust, and because it is published at a high velocity on trending topics, the algorithm is tricked into thinking it is fresh, relevant news. The content begins to appear in users’ Discover feeds, often with sensational headlines designed to maximize clicks.
Step 4: Monetization
Every click on these spam articles generates advertising revenue for the spammers. With the massive reach of Google Discover, even a low-quality site can generate significant income before it is eventually caught and de-indexed. In France alone, one data journalist has tracked over 8,300 AI-generated sites attempting to use this exact “cash machine” tactic.
Why Google’s Defenses Are Failing
Google’s official stance is that its “robust spam-fighting systems” keep the “vast majority of spam out of Discover.” However, the evidence from users’ feeds tells a different story. The current crisis exposes several critical weaknesses in Google’s defenses.
- The Trust Loophole: Google’s algorithms have historically placed a high degree of trust in older, established domains. Spammers have weaponized this trust by buying these expired domains, creating a loophole that allows their low-quality content to bypass the normal quality filters applied to new sites. While Google has policies against this, their automated systems are clearly struggling to enforce them at scale.
- The AI Content Detection Problem: Despite claims to the contrary, Google’s systems are still not able to reliably distinguish between human-written and AI-generated content, especially when the AI content is designed to mimic human writing styles. As Google’s own John Mueller has stated in the past, AI-generated content intended to manipulate search rankings is considered spam, but detection remains a massive challenge.
- The Scale of the Attack: The sheer volume of AI-generated content being produced is overwhelming Google’s systems. Spammers can now create thousands of articles in the time it used to take to write one, making manual review impossible and putting immense strain on automated systems like SpamBrain, Google’s AI-powered anti-spam system.
The Promised Fix: What Can Google Actually Do?
In its statement, Google promised it is “actively working on a fix that will better address the specific type of spam that’s being referenced here.” While the details are not public, we can infer the likely components of this fix based on the nature of the problem.
Potential Elements of the Fix:
- De-Trusting Expired Domains: The most effective and necessary step would be to implement a “trust reset” for expired domains. When a domain changes ownership, its previous authority score should be significantly reduced or completely wiped, forcing it to earn trust from scratch like any new domain. This would close the primary loophole that spammers are exploiting.
- Enhanced AI Content Demotion: Google’s SpamBrain will likely be updated with more sophisticated models specifically trained to identify the patterns of low-quality, AI-generated content, even if it’s grammatically correct. This could involve analyzing the content for factual accuracy, originality, and the “hallucinations” common in AI-written text.
- Stricter Publisher Vetting for Discover: To maintain a “high bar for quality,” Google may need to move away from a purely algorithmic approach to Discover eligibility. They may re-introduce a level of manual review or implement stricter vetting criteria for which sites are allowed to appear in the feed, prioritizing established, trustworthy news organizations over new or repurposed domains.
The Broader Implications: A Crisis of Trust
The AI spam problem in Google Discover is more than just a technical issue; it’s a crisis of trust that has implications for the entire digital information ecosystem.
- For Users: The flood of fake news and low-quality content erodes trust in Google as a reliable source of information. It forces users to be constantly vigilant and skeptical, degrading the overall user experience.
- For Legitimate Publishers: For years, publishers have invested heavily in creating high-quality, original content to serve their audiences, often relying on Discover for a significant portion of their traffic. Seeing that traffic evaporate overnight, replaced by AI-generated spam, is a devastating blow to their business model and a disincentive to continue investing in quality journalism.
- For Google: This crisis strikes at the core of Google’s mission “to organize the world’s information and make it universally accessible and useful.” If the information being organized is garbage, the entire value proposition collapses.
Conclusion: The Clock Is Ticking
Google’s acknowledgment of the AI spam problem is a welcome first step, but the clock is ticking. Every day that this loophole remains open, more low-quality content pollutes the information ecosystem, more users are misled, and more legitimate publishers are harmed. The promised “fix” cannot be a simple patch; it must be a decisive and fundamental change in how Google identifies, vets, and amplifies content in the age of AI. The future of trustworthy information discovery depends on it.
