In rare move, FTC sets aside Rytr Order for burdening AI innovation (and failing to plead violations)

Background

Last year, the FTC filed a complaint against Rytr, which currently markets itself as an “AI writing assistant for short-form content,” offering over 40 use cases ranging from emails and blog posts to social media ads and customer testimonials through both free and subscription tiers.

The FTC took particular issue with Rytr’s “Testimonial & Review” use case, alleging that the tool could “quickly generate an unlimited number of detailed and genuine-sounding reviews with minimal input,” and that some subscribers had generated tens of thousands of reviews –including in a single month. According to the FTC, the “Testimonial & Review” tool’s “likely only use” was to facilitate “fake reviews” to deceive customers and therefore had “no or de minimis legitimate use.” The FTC alleged two Section 5 violations for deception and unfairness and, in December 2024, Rytr entered a consent order (“Rytr Order”) that, among other obligations, banned it from offering or promoting any review or testimonial generation service.

Two Republican Commissioners dissented, laying the groundwork for the December 2025 set-aside.

• Now-Chair Andrew Ferguson warned that treating a generative AI tool as categorically illegal “merely because of the possibility that someone might use it for fraud” risks turning “honest innovators into lawbreakers” and “strangling a potentially revolutionary technology in its cradle.”
• Then-Commissioner Melissa Holyoak emphasized that “[u]nfairness requires proof – not speculation – of harm,” and cautioned against banning products with useful and lawful features simply because they could potentially be misused. She objected against that enforcement approach as inconsistent with both the Commission’s unfairness authority and its aim to promote innovation.

One month later, President Trump issued Executive Order (EO) 14179, “Removing Barriers to American Leadership in Artificial Intelligence” and announced his administration would pursue an “AI Action Plan.” That plan, released earlier this year, requires the FTC to, among other things, “review all FTC final orders, consent decrees, and injunctions, and, where appropriate, seek to modify or set-aside any that unduly burden AI innovation.”

Rationale to reopen and vacate the Order

The two current members at the typically five-member Commission – Chair Andrew Ferguson and Commissioner Mark Meador, both Republicans – announced that the agency had reviewed the Rytr Order pursuant to the AI Action Plan and determined that the matter should be re-opened and set aside because it was against the public interest. They concluded that the Commission had not sufficiently pled any violation of Section 5 and that it unduly burdened AI innovation in violation of EO 14179 and the AI Action Plan. Rytr consented to vacating the order.

The set-aside focuses on the two theories of liability alleged in the initial matter: (1) “means and instrumentalities” of deception; and (2) unfairness.

(1) Means and instrumentalities

The original complaint alleged that Rytr violated Section 5 of the FTC Act by providing its users and subscribers with the “means and instrumentalities to deceive” others by generating deceptive consumer reviews.

In vacating the Rytr Order, the current FTC leadership found the complaint did not allege conduct that met any of the traditional tests for “means and instrumentalities”-based liability. It did not allege Rytr made any deceptive statements; it did not allege the tool was inherently deceptive; and it did not allege Rytr knew or had reason to know that its tool was being used for deceptive purposes.

The Commission underscored that the mere potential for misuse does not establish a Section 5 violation. Echoing Chair Ferguson’s earlier dissent, the FTC warned that expanding the means and instrumentalities theory as applied in Rytr to tools with both lawful and unlawful uses could “condemn anyone who makes ‘pencils, paper, printers, computers, smartphones, word processors, typewriters, posterboard, televisions, billboards,’ or any other tool that someone somewhere might use to create or disseminate a false advertisement.”

(2) Unfairness

The previous Commission alleged Rytr’s conduct met the Section 5 definition of unfairness primarily because the AI-generated customer review use case had “no or de minimis legitimate use.”

Current FTC leadership rejected that conclusion, finding that “almost all tools have both legal and illegal uses” and the very features that raised concern for the previous Commission – that is, the ease with which one could generate a customer review with minimal user effort – was, as then-Commissioner Holyoak described in her prior dissent, “a feature, not a bug.” Current FTC leadership also emphasized that an unfairness theory requires adequately alleging that the AI platform causes or is likely to cause substantial injury to consumers, and found that there was neither empirical nor qualitative evidence to support such a showing in the original Rytr Order.

Takeaways

• A rare reversal may signal a meaningful shift. Re-opening and setting aside an FTC consent order is an unusual step, and even more rare when the FTC does so without a petition from a consent order recipient. In the past 25 years, the FTC has only taken the step of re-opening a decision on its own accord a handful of times – and nearly always to modify the order, not to vacate it. We will continue to monitor whether this is a signal that the current, and potentially future, Commissions may be more willing to revisit the decisions of their predecessors.
• Tools with both lawful and potentially unlawful uses are not inherently unlawful. This decision shows the current FTC’s view of AI tools: making life easier is a benefit to consumers, even if the benefit is cutting down the time and effort needed to write a review or testimonial. The FTC will not assume that a tool that could potentially be used for improper purposes is itself improper, especially where, as here, there is little to no evidence of any actual injury to consumers nor likelihood of injury.
• Review integrity remains top of mind for the FTC. Review integrity has been a major focus of the FTC in recent years and this decision should not be taken as an indication that the FTC shifting that focus. On the same day it issued this order, the FTC issued warning letters to 10 companies regarding potential violations of the FTC’s Consumer Review Rule.
• The FTC remains focused on AI, particularly AI safety. The current Commission has issued statements showing it continues to prioritize AI and in particular, AI safety related issues. Notably, it is doing a broad investigation of advertising, safety, and data handling practices for various AI companies under its Section 6(b) authority.
• Expect continued recalibration on AI, and possibly a shift to state-level enforcement. The set-aside suggests the FTC will continue to refine its approach to enforcement regarding AI, balancing innovation with consumer protection and grounding actions in evidence rather than speculation. Meanwhile, enforcement actions may become more common at the state level, particularly if state regulators seek to address changes in federal enforcement priorities through analogous state laws.

Authored by Lance Murashige, Sophie Baum, Harsimar Dhanoa, and Soojin Jeong.