• by Wendy Davis , December 18, 2024

Rejecting arguments by a tech industry group, the Federal Trade Commission on Wednesday said it had finalized a settlement with Rytr that prohibits the artificial intelligence company from advertising or selling a service that generates reviews.

The deal resolves allegations that Rytr enabled subscribers to its service to generate fake reviews that would deceive consumers.

The organization TechNet, which counts Amazon, Google and Meta among its members, opposed the settlement, arguing in comments submitted last month that the FTC's enforcement action was “overbroad” and could hinder innovation in artificial intelligence.

The group argued that there were no allegations that Rytr itself made misleading statements, or knew its service was being used to create fake reviews. TechNet added that treating generative artificial intelligence as “categorically deceptive” would create a precedent that could result in other companies -- such as those offering word processors or graphics editors -- being held liable for enabling deceptive practices by their customers.

The FTC countered this week in a letter to TechNet that the enforcement action against Rytr was “narrowly focused” on the facts -- including that 24 subscribers allegedly generated more than 10,000 reviews each, and that one subscriber generated reivews for garage door repair companies throughout the country.

TechNet's arguments against the settlement was similar to objections raised by the FTC's two Republican commissioners who said in September they disagreed with the decision to prosecute Rytr.

“Treating as categorically illegal a generative AI tool merely because of the possibility that someone might use it for fraud is inconsistent with our precedents and common sense,” Andrew Ferguson, who will take over as head of the agency next month, said in a written dissent. “And it threatens to turn honest innovators into lawbreakers and risks strangling a potentially revolutionary technology in its cradle.”

The advocacy group Consumer Reports had urged the full FTC to reject the Republicans' position, arguing that Rytr “created a platform whose overwhelming use case was to be used for fraud.”

“Rytr’s interface contained few fields for users to input details about the products or services they wanted to review; instead, the company offered various toggles for desired tone or levels of creativity,” Consumer Reports said in written comments. “Based on these limited inputs, Rytr would generate detailed reviews with specific invented anecdotes unrelated to any information provided by the customer. Rytr set no limits on how many reviews a customer could create, or even how many reviews a customer could create for one particular company.”