Nirav Tolia co-founded Epinions in 1999, one of the first user-generated review sites, on the premise that the collective judgment of the masses would sum to something useful. The company was later sold to eBay. Tolia writes that the industry it helped launch "would eventually be gamed," mixing genuine experiences with ratings that can be manufactured or bought.
In 2024, the Federal Trade Commission finalized a rule making deceptive review and testimonial practices illegal. Tolia, now Nextdoor's CEO, writes that the platforms at the center of the problem "aren't oblivious to it": they have built out enforcement teams, sued fake-review brokers, and removed hundreds of millions of fraudulent reviews over the past year alone. Even so, Tolia writes, the fake and manipulated share of reviews "kept growing anyway."
According to Tolia, 85% of consumers now suspect online reviews are fake, and 88% say they can't tell which ones are real. He writes that the habit of relying on reviews hasn't died, but trust in them has eroded.
Tolia's argument, as he lays it out, is that any trust signal is only as durable as what it costs to fake and what a person loses when caught faking it. Star ratings, he writes, fail both tests: no one knows who is behind an anonymous rating, and no one pays a price when it's exposed. A named recommendation from a neighbor, by contrast, carries reputational risk if it steers someone wrong — what Tolia calls "skin in the game."
Early research fielded this summer by the consumer-insights platform Echo, cited by Tolia, found that nearly half of respondents had suspected a fake or manipulated review in the past year, and 81% said they would trust a recommendation more if the person making it were held socially accountable.
Tolia also writes that artificial intelligence complicates the picture further. He cites a 2025 study from Nottingham University Business School finding that AI-generated fake reviews are now indistinguishable from real ones, to both human readers and other AI models. He writes that businesses have no way to influence the answers large language models produce about them, and that some have been targeted by "AI bombing" — the use of generative AI to flood a business with large volumes of fake or misleading reviews.
What Tolia describes is a federal rule that made deceptive reviews illegal on paper, enforcement teams and lawsuits from major platforms, and a fraud problem that grew regardless. That sequence is worth sitting with for anyone who wants to know what regulation can and cannot buy.
The response Tolia proposes isn't another rulebook — it's a product built on real identity, where reputational cost does the enforcing that an anonymous rating never could. Whether or not that model scales, it's a reminder that the FTC's ban addressed the label on the box, not the incentive to lie in the first place. For an economy increasingly mediated by AI-generated content of uncertain origin, that gap is the one worth watching.
