New Addiction Lawsuits May Be Best Bet to Tame Big AI

A jury just found Meta and YouTube guilty of knowingly programming its products to cause social media addiction.¹ It awarded $6 million in compensatory damages for causing psychiatric problems & million in punitive damages for being irresponsible. In another New Mexico case, Meta was found guilty of violating state law requiring it to safeguard users from child predators and ordered to pay $375 million punitive damages.² Dozens of other cases are pending. This is encouraging evidence supporting the hope that, like Big Tobacco and Big Pharma, Big AI can be tamed via lawsuit.

In the 1990s, Big Tobacco seemed too big and too powerful to fail. But a $200 billion legal judgment and accompanying increased regulation cut Big Tobacco down to size.³ Cigarette smoking in the US adult population dropped from 43% to 12%, the stock valuation of tobacco companies has since badly trailed the stock market, and Big Tobacco lost its ability to buy politicians or co-opt regulators.⁴ In the same vein, the Sackler family seemed immune to punishment and regulation as it collected billions of dollars addicting millions of Americans to powerful medical opioids. But a successful lawsuit bankrupted their company and is clawing back $7.4 billion of their own ill-gotten gains.⁵

Notable in both these cases was the government's inability to regulate the addictions consciously caused by greedy and unscrupulous companies. The method to reduce the public health harm was clear, but the political motivation was too weak to withstand the huge lobbying power and campaign contributions mobilized by Big Tobacco and Big Pharma.

As Mark Twain put it: "History doesn’t repeat itself, but it sure does rhyme." By executive orders, Trump has blocked all federal and local regulation on AI. The fate of our country, and future of our species, are now in the hands of the handful of greedy, power hungry Big AI companies. As we discussed in a previous piece, class action lawsuits that inflict an enormous financial and reputational hit may be the only way to contain the public health, employment, environmental, political, and financial harms caused by artificial intelligence.⁶

There is good news: a new series of lawsuits are about to be adjudicated that have striking parallels to those which contained Big Tobacco and Big Pharma.⁷ Dov Grunschlag, Esq, a legal expert, will bring us up to speed on the legal questions, how they may play out, and how this may impact the negative role of chatbots on mental health.

Allen Frances, MD: What are the details of these upcoming legal cases involving AI?

Dov Grunschlag, Esq: The common claim in these lawsuits is that social media platforms are addictive—not because of their inherent nature, as tobacco and opioids are, but worse by design. As a result, the consumers at which the content is aimed—teens and younger—suffer potentially lifelong damage in the form of anxiety, depression, eating disorders, self-harm, and in some cases, suicide.

Frances: What will be plaintiffs' claims?

Grunschlag: The focus is on design features (eg, prioritizing engagement, infinite scroll, auto video play, algorithmic recommendations) that hook users into social media addiction. The underlying legal claim is social media platforms create a “product”—like tobacco, like opioids—and that its manufacturers and distributors should be responsible for the harm it causes, on several legal theories: a violation of a duty of care, perhaps a heightened duty owed to young people, or “strict liability,” meaning liability regardless of “fault” in an ordinary sense.

Frances: What is the likely defense rebuttal?

Grunschlag: The defense is likely to rest on factual and legal grounds. It will dispute that as a general proposition, exposure to and use of social media cause addiction, as distinguished from personal choice. It will attempt to scrutinize individual plaintiffs for other causes of harm suffered. It will reject the analogy to “product,” arguing that it is not responsible for content generated by others, claiming the protection of a federal law (Section 230 of the Communications Decency Act), and the First Amendment (free speech). These legal issues have not yet been resolved in definitive fashion.

Frances: What do you anticipate expert testimony will be like?

Grunschlag: There will likely be testimony from mental health experts and technology experts, on both sides. Mental health experts for the defense will likely review relevant research on the addictive qualities of social media and claim that defendants prioritized "engagement" programming despite knowing it would escalate the risk of addiction in users. They will support the general claim with examinations of individual plaintiffs to demonstrate how addiction occurred and its harmful consequences.

Mental health experts for the defense will attempt to narrow the definition of addiction, emphasize preexisting problems, support personal choice, and describe positive effects of social networking. Technology experts for the plaintiffs will describe how programming decisions negatively impact user behavior. Technology experts for the defense will emphasize safety features built into the programming and low rates of problem use.

Frances: Given the high stakes, how can the companies let these cases go to juries, rather than be settled out of court?

Grunschlag: Several lawsuits have been settled before trial—eg, TikTok settled a case in Los Angeles on the eve of trial—avoiding the risk of a worse result. That is probably the option preferred by both sides. But a case may go to trial if the plaintiff in a given case insists on a payment that the defense believes is substantially higher than its worst-case scenario at trial—a risk/reward analysis notoriously difficult in uncharted waters. There is a related impetus behind taking a case to trial: to gain some knowledge about how juries respond to these cases, and to resolve the unsettled legal issues, which will in turn influence and shape how consumers, social media companies, large insurance companies, and potentially government regulators deal with these matters going forward.

Frances: Do you think these lawsuits could have a significant impact on Big AI regulation and behavior?

Grunschlag: While the consumer/user interaction with social media is not the same as that with AI directly, the outcome of the lawsuits against social media companies could have a significant impact. How juries react in these cases could well be predictive of how they will react to cases against AI companies. The legal theories that gain traction in these cases could serve as precedents. What the litigation teaches about the challenged design features of social media could have a bearing on how bots are programmed. Technologists should, and likely will want to collaborate with mental health experts on how best to mitigate potential adverse consequences. In both areas, the litigation, regardless of outcome, could generate political pressure for nationwide regulation.

Concluding Thoughts

It is terrifying that Big AI companies have virtually no external constraints and few internal scruples. As with Big Tobacco and Big Pharma, legal recourse is likely to be the most powerful force for reform. Media shaming and intense advocacy for government regulation must be pushed to their limits, but it would be a triumph of hope over experience to expect much reform of company behavior to result from them. Successful lawsuits against Big AI will likely be the most powerful way to induce it to build in protections for our patients.

Dr Frances is professor and chair emeritus in the department of psychiatry at Duke University.

Mr Grunschlag is a lawyer in labor and employment law, with experience in class action lawsuits.

The opinions expressed are those of the author and do not necessarily reflect the opinions of Psychiatric Times.

References

1. Hays K, Saad N, Morris R. Campaigners welcome Meta and YouTube’s defeat in landmark social media addiction trial. BBC. March 27, 2026. Accessed March 30, 2026. https://www.bbc.com/news/articles/c747x7gz249o

2. Hays K. Meta told to pay $375m for misleading users over child safety. BBC. March 24, 2026. Accessed March 30, 2026. https://www.bbc.com/news/articles/cql75dn07n2o

3. Master settlement agreement. Office of the Attorney General. Accessed March 18, 2026. https://oag.ca.gov/tobacco/msa

4. Trends in cigarette smoking. American Lung Association. Accessed March 18, 2026. https://www.lung.org/research/trends-in-lung-disease/tobacco-trends-brief/overall-smoking-trends

5. Mann B, Martinez A. Purdue Pharma and Sackler family members to pay $7.4B in national opioid settlement. NPR. January 24, 2025. Accessed March 18, 2026. https://www.npr.org/2025/01/24/nx-s1-5272769/purdue-pharma-and-sackler-family-members-to-pay-7-4b-in-national-opioid-settlement.

6. Frances A, Grunschlag D. Can a class action lawsuit force big AI to make chatbots safer? Psychiatric Times. October 21, 2025. https://www.psychiatrictimes.com/view/can-a-class-action-lawsuit-force-big-ai-to-make-chatbots-safer

7. Panettieri J. Generative AI lawsuits timeline: legal cases vs OpenAI, Microsoft, Anthropic, Google, Nvidia, Perplexity, Salesforce, Apple, and more. Sustainable Tech Partner. February 26, 2026. Accessed March 18, 2026. https://sustainabletechpartner.com/topics/ai/generative-ai-lawsuit-timeline/