Rewriting Rytr: The FTC sets aside a Final Order to implement America’s AI Action Plan and limits means and instrumentalities liability

On December 22, the Federal Trade Commission (FTC) reopened and set aside a 2024 administrative order against Rytr LLC. There’s a lot to unpack in the six-page set-aside order, which states that the case against Rytr unduly burdened artificial intelligence (AI) innovation in violation of the Trump Administration’s Artificial Intelligence Executive Order and America’s AI Action Plan. Here’s what to know.

The Rytr matter

In September 2024, the FTC announced law enforcement actions against five companies it alleged “relied on artificial intelligence as a way to supercharge deceptive or unfair conduct that harms consumers.” It dubbed its sweep, “Operation AI Comply.”

At the time, FTC staff noted that three of the five cases involved “oldies but not-so-goodies: deceptive business opportunity scams that claim to use AI to help people earn more money, faster. The other two, however, were different. One of those was the case against Rytr.

According to the FTC’s complaint, Rytr sold an AI-enabled writing assistant with a tool that allowed customers to generate unlimited online reviews and testimonials based on limited, generic input. The complaint claimed that offering this service, which subscribers could use to produce thousands of fake reviews, constituted an unfair trade practice. Additionally, according to the complaint, by offering this tool, Rytr gave its customers the “means and instrumentalities” to deceive people.

Then-Commissioner Melissa Holyoak and then-Commissioner, now Chairman Andrew Ferguson didn’t like the action at the time. They voted against it and issued a joint dissent, raising two categories of concerns relating to sufficiency of the complaint.

First, they asked whether the allegations of the complaint met the FTC Act’s three-pronged test for an unfair business practice, which provides that an act or practice is unfair if it causes or is likely to cause substantial injury to consumers not reasonably avoidable by those consumers, and is not outweighed by countervailing benefits to consumers or to competition. The dissent argued that the conduct alleged in the Commission’s complaint did not meet this test. Specifically, the dissenters noted that the complaint focused on potential injury to consumers without alleging Rytr’s conduct actually injured consumers. Moreover, the dissenters said, the complaint did not appropriately weigh the countervailing benefits Rytr’s service offered to consumers and competition. Indeed, they noted, the complaint seemed to acknowledge that Rytr’s drafting aid and suite of products offered real benefits, including helping people “save time and achieve their goals with less work."

Second, the dissenters argued against the Commission’s use of “means and instrumentalities” liability, explaining that application of the theory typically involves a scenario where a marketer publishes a deceptive statement and places it in the hands of others, thereby giving those others the means to deceive consumers. In this case, however, even if reviews drafted with Rytr contained misrepresentations – which was unclear from the complaint – the complaint contained no allegation that Rytr itself made material misrepresentations. Accordingly, the dissenters said, means and instrumentalities liability was not appropriate.

America’s AI Action Plan

As we’ve previously discussed, President Trump’s January 23 Executive Order issuing America’s AI Action Plan includes some very specific asks of the FTC. In furtherance of AI innovation, the EO instructs the agency to:

1. Review all investigations initiated under the Biden administration to confirm they “do not advance theories of liability that unduly burden AI innovation”; and
2. Review all orders and “seek to modify or set-aside any that unduly burden AI innovation.”

With the December 22 announcement, the FTC is taking a clear and public-facing step to carry out those directives.

Order Reopening and Setting Aside the Rytr Order

Given the Republican commissioners’ (including now-Chairman Ferguson’s) no-votes and detailed dissenting statement, the Rytr matter is an obvious choice to set aside as “unduly burden[ing] AI innovation.” Thus, the order implementing the set-aside first recaps the procedural history, then includes the unsurprising fact that Rytr has consented to the set-aside, and wraps up by outlining many of the same points set forth in the original dissent in a section titled, “The order is contrary to the public interest.”

Like the dissent, the set-aside order focuses on whether the underlying complaint pled sufficient facts to meet the tests for an unfair business practice and means and instrumentalities liability. When it comes to unfairness, the argument is more or less the same as the underlying dissent, with the addition of a pronouncement FTC staff will likely hear over and over in negotiations to come: “When a product or service has a pro-consumer, legitimate use, it is plainly wrong to allege the product is inherently deceptive.”

However, the set-aside order also provides a much more detailed analysis of means and instrumentalities liability, and suggests that marketers can expect a more constrained application of that theory moving forward. Specifically, the set-aside order argues that the FTC inappropriately applied means and instrumentalities in the Rytr case because the complaint failed to alleged facts fitting any of the situations where courts have found such liability appropriate or “surface any other cognizable application of the theory.”

The three scenarios observed to be acceptable uses of means and instrumentalities liability are:

1. When suppliers provide deceptive marketing material to sellers, who use that material to deceive consumers;
2. When an entity provides others with a product or service that is inherently deceptive; and
3. When the defendant knows, or has reason to know, that the person to whom a product or service was supplied will use it to violate Section 5.

By listing these three use cases, the Commission has both strongly signaled it will be skeptical of means and instrumentalities liability in other cases, and provided a clear roadmap to argue against application of the theory.

So, what’s the bottom line?

There’s plenty to discuss here, but we see a few immediate takeaways for businesses and practitioners:

• The FTC’s carrying out its orders from America's AI Action Plan. The FTC is doing what the President asked it to do: identifying and rescinding orders it thinks burden AI innovation. Companies subject to FTC orders or undergoing investigation regarding products or services with a nexus to AI should look closely at the set-aside order and assess whether their matters align with it.

• Expect closer scrutiny of means and instrumentalities liability. Over time, the Commission’s use of means and instrumentalities liability has waxed and waned. It looks like we may be moving into a down cycle. If means and instrumentalities comes up in an investigation, look at the set-aside order and consider whether the allegations in the case fall into one of the three categories the FTC outlined.

• Watch for continued deference to “pro-consumer, legitimate use[s].” As Chairman Ferguson has said in other contexts, the Trump-Vance FTC intends to exercise its law enforcement functions where appropriate, and then “get out of the way” for legitimate business. We’re seeing that philosophy in action here.
• The FTC’s still cracking down on fake reviews. We’d be remiss if we didn’t acknowledge that, approximately 30 minutes after announcing it was setting aside the Rytr order, the Commission announced it had sent 10 companies warning letters about potential violations of the Consumer Review Rule. Whether this timing was intentional is unclear, but, regardless, the announcement clarifies that business should not construe setting aside the Rytr order as a sign the Commission’s going to ease up on fake reviews.