August 19, 2026, 23:52 GMT | Insight When more than 40 US state attorneys general sued Meta Platforms on addictive design claims in 2023, a senior Meta user experience researcher called the AGs' allegations "somewhat hard to refute" in an internal chat, a jury in Tennessee heard Wednesday.

As the State of Tennessee moved toward completion of its case in its addictive-design trial against Meta that began July 27 in Nashville (see here), Thomas Wynter, the senior UX researcher, said in a video deposition that within the company, he was a critic of Meta's legal and communications strategy.

While Wynter testified that he believed in the company's efforts to reduce the prevalence of harmful content on Instagram and Facebook, and said Meta had made surprising and significant progress in that effort in 2024, he had blasted in the 2023 internal chat Meta's public defense to the state lawsuits — that it had developed more than 30 safety and family-support tools to limit minors' excessive use of its social media apps.

"I have to ask: Are the compliance people reading the same litigation and legislation that I am?" Wynter wrote in one internal post to a colleague viewed by the jury, complaining that the safety tools Meta was citing in its reply to the states' complaints were a "mismatch," and that the tools were "concerned with optimizing for engagement."

"So really, the algorithm is absolutely tuned to maximize engagement in an maximally empirical, principle-less way. I am not sure that the youth legal people realize this, and that it applies to teens just as much as [the general population of users]," Wynter wrote in one of his 2023 chats.

In other internal comments about news reports regarding the more than 40 states that sued in October of that year over the social media platforms' allegedly addictive design (see here), Wynter wrote in an internal chat that "the things contained in the [attorneys general's] lawsuits are somewhat hard to refute."

Following an objection by Meta, Russell T. Perkins, the court chancellor overseeing the trial, issued a limiting instruction to the jury that Wynter is not a lawyer, and the jurors should not interpret that statement as a legal judgment.

The jury saw several of Wynter's internal chats with other Meta researchers that appeared to back claims by Tennessee and other states that Instagram and Facebook's features were knowingly designed to maximize the amount of time that users, particularly teens, spent on Meta's apps.

In another internal chat from 2022 seen by the jury, Wynter noted that "nobody has noticed" Meta's algorithms were steering problematic content to teen users, telling a colleague in a profanity-laden chat message, "how much am I gonna care that deprivation studies show wellbeing gains from not using our s****y products."

Wynter acknowledged under questioning from the state that the product he was referring to was Facebook.

"I feel like I'm in a tobacco company strategy meeting," Wynter complained in yet another internal Meta chat the jury was shown.

In his video deposition, Wynter struggled to explain the tobacco company reference, saying "I think really what I was trying to do is guilt my manager into getting the engineers" that the UX researcher needed to study ways to reduce the prevalence of harmful content on Meta's apps.

The states, along with a child psychologist who testified as an expert witness for Tennessee on Wednesday, have explicitly analogized Meta's alleged behavior to large tobacco companies that failed to disclose their own research showing their products were harmful, particularly for minors (see here).

— Design features harms —

With the state having wrapped up its testimony from senior Meta executives including Instagram chief Adam Mosseri, the jury heard from what lawyers for Tennessee said would be their last witness: child psychologist Parker Huston.

One "opinion that I came to is that some of Instagram's features, like push notifications, autoplay, infinite scroll, and ephemerality or temporary content, encourages behavior that shares characteristics with behavioral addiction," Huston told the jury, discussing the expert report he authored.

Those design features, Huston told the jury, "contribute significantly" to the 15 to 20 percent of the teen patients he treats in his practice in Ohio who have conditions caused by social media use. "There's a contributory causal relationship between those design features and some of the behaviors that we're talking about related to addiction," Huston said.

Had Meta made public some of its internal studies, his clinical practice would have been improved in dealing with patients whose mental health was harmed by excessive social media use, he said. "I think it would have helped me have a little bit more confidence that a company like Meta is invested in making product improvements that actually benefit their users."

On cross-examination from Meta lawyer Kevin D. Horvitz, however, Huston acknowledged that some use of social media could be beneficial to teens. "I think it could be beneficial for a variety of reasons in certain circumstances," he said.

But those benefits — a teen finding like-minded people, being allowed to be their authentic self and express themselves freely — don't require design features that encourage excessive use, Huston said in response to a question from Horvitz.

"We don't need ephemerality for that. We don't need infinite scroll, or autoplay. We don't need those things to achieve those ends," said Huston, who acknowledged he had been paid about $40,000 by the state for his work.

— DM problems —

Late Tuesday, the jury heard in Mosseri's direct video testimony that sexual harassment toward minors on Instagram through direct messages was a problem for years before Meta was able to develop a nudity filter for content sent through DMs.

In 2018, Guy Rosen, Meta's chief information security officer, emailed Mosseri about how direct messages on Instagram are "where all the bad stuff happens," according to a note seen by the jury.

"From the 'less bad' tier 2 sexual harassment like dudes sending [photos of male genitalia] to everyone; to the tier 1 cases where they end up doing horrible damage ... I do think the safety cases are much more important on FB+IG than on WA because FB+IG are a directory," Rosen wrote, referring to Facebook, Instagram and WhatsApp.

"My whole question that I kicked out this email for was trying to understand the state of our safety tooling in the context of Instagram messaging," Mosseri said.

When pressed on what "tier 1" sexual harassment would be, Mosseri could not name a specific example, "but it sounds pretty horrific," he said.

The jury also heard a 2020 podcast in which Mosseri discussed where people shared more intimate images, which he said is in messages.

Asked whether the images Rosen described were actively being monitored by Meta, Mosseri said "Not necessarily."

"So we do actually scan direct messages for known instances of child-exploited imagery, but we do not scan direct messages or private direct messages for everything. We're trying to balance people's interests in privacy, not wanting us to scan everything, with our interests in safety," he said.

Meta has also since implemented a nudity control feature to ensure people don't immediately see that content in their messages, Mosseri said, noting the feature was implemented in 2024.

"Earlier, you were trying to make it sound like you were getting ramped up to protect kids from getting [photos of male genitalia] on Instagram direct messaging, but it didn't happen until April of 2024," the lawyer for the state said.

Please email editors@mlex.com to contact the editorial staff regarding this story, or to submit the names of lawyers and advisers.