Class Actions in the Age of AI

Commentary

Class actions have played a major role in transforming society since the modern class action was born in 1966. That was when the Federal Rules of Civil Procedure in the United States were amended to create the “opt-out” mechanism.

Until then, class actions had required individuals to proactively sign up for a class action lawsuit and opt into it. That proved to be a major impediment to disadvantaged groups. The poor, the marginalized, and the legally or otherwise illiterate had to know that a class action on their behalf had even been started and then seek out the lawyers handling it. In addition, they might fear retaliation from the powerful interests standing in opposition to the claims.

Those concerns were eliminated by the invention of the opt-out mechanism. It automatically includes within a class action any individual who fits the definition of the class as defined by the lawyers handling the case and as certified by a judge. It takes an active step to opt out of the litigation and is fully inclusive of everyone within the class.

This revolutionary change increased the clout of class action litigation exponentially. True, there had been past successes, most notably the 1954 case of Brown v. Board of Education, a consolidation of five separate class actions, in which the U.S. Supreme Court ruled against the segregation of students in school along racial lines. But the true power of the class action was yet to be unleashed until this simple and creative procedural amendment was made.

Canada was slow to follow. Quebec was the first to adopt similar class action laws in 1978, followed by the other provinces, and Ontario specifically in 1992. The globalization of the class action tool continues, with the EU most recently having adopted the 2020 Representative Actions Directive, which requires all member states to implement consumer class action legislation.

Class actions have provided compensation and precipitated legal reform across a vast array of harms, from corporate misfeasance to public policy transgressions.

Asbestos is no longer used in the construction of homes, smoking tobacco in public places has been banned, and Ford Pintos no longer explode on impact, largely as a result of class action litigation. A multibillion-dollar fund was established for indigenous peoples who suffered abuse in residential schools.

The Canadian blood system was reshaped by tainted blood victims who had contracted HIV and Hepatitis C from blood transfusions. The structural changes that stripped the blood program from the Canadian Red Cross and transferred it to Canadian Blood Services and Hema Quebec arose in part both from 13 class actions and the conclusions of the Krever Public Inquiry, but the compensation packages put together for the blood victims by the provinces came about primarily because of the class action lawsuits.

Class action litigation, however, cannot replace appropriate public policy. Only public policy can prevent social harms. Litigation generally propels reform and compensation only after harm has been suffered. The structural reforms are negotiated privately by lawyers and lack the intense public debate that accompanies legislative reform, possibly placing the broader public interest at risk in the name of the almighty dollar. Still, the class action has proven its worth as a valuable tool that provides access to justice and acts as a conduit for social and legal reform.

As we grapple with the possible social fallout from social media and AI, what does the future litigation landscape hold for these two pillars of the post-modern world?

The future has already arrived. Social media and AI class action litigation are already proliferating at a rapid pace. And this is happening even before AI has matured to its full potential and, as some would say, to its complete destructive and destabilizing power.

With the USA and China engaged in a race for AI supremacy—with smaller powers now seeking greater autonomy in the AI world as the two hegemons increasingly use their heft to impose their worldviews on others—the prospects for international cooperation in regulating AI currently seem distant. The lobbying grip of Big Tech on some governments seems tighter than ever and makes the prospects for regulation even dimmer.

In the meantime, the recent Meta class action settlement initiated by several U.S. states, to the tune of US$18 billion for harms caused to children by Facebook, is only the beginning.

Canadians have started similar class litigation against Meta, most notably in B.C.

In another suit, the Toronto School Board sued ByteDance (TikTok), Snap Inc., Meta, and others on behalf of its 23,800 students.

Although technically not a class action, the school board argues that a board is the statutory representative of its student population and that the Big Tech firms have negligently interfered with the board’s statutory mandate to promote student achievement and well-being. In March 2025, the Ontario Superior Court denied the defendants’ motion to strike the claim and held that the plaintiffs had reasonable arguments to advance that should go to trial.

In the USA, Anthropic, the designer of Claude, recently agreed to a US$1.5 billion settlement with authors for breach of copyright after the class action was certified by a California judge. The judge also held that the defence of fair use did not apply to the copyright action under the circumstances. More such litigation is expected.

Similar class actions have recently been brought by authors and publishers against Google, now a subsidiary of Alphabet Inc., Microsoft, and Apple, while musicians and composers have begun class actions against the purveyors of AI-generated music.

As AI continues to develop unhindered by meaningfully regulatory frameworks of any kind, other than privacy protection, new harms will likely arise and the impact on the job market may be profound. Yet the job prospects for class action litigation lawyers look very bright indeed, assuming, of course, that those who see AI as an existential threat to humanity prove to be wrong.

Views expressed in this article are opinions of the author and do not necessarily reflect the views of The Epoch Times.