A new wave of generative AI class actions is now building directly on the litigation industry's early courtroom successes, according to an analysis published by the law firm Holland & Knight. For photographers, stock agencies and licensing businesses, the trend signals that courts are increasingly willing to let copyright and right-of-publicity claims against AI developers proceed past the motion-to-dismiss stage.

The report, titled "Recent GenAI Class Actions Build on Early Successes and Break New Ground," examines how plaintiffs' attorneys are refining their strategies after initial rulings in earlier cases. Where first-generation complaints often collapsed on technical grounds, the newer filings borrow from complaints that survived early dismissal — and extend the theories into areas the courts have not yet addressed.

Why does this matter for photo businesses?

Photography and visual-art licensing sit at the center of the GenAI copyright disputes. Training-data lawsuits allege that AI developers scraped copyrighted images at scale without permission or payment. If courts accept those theories, the litigation could reshape how image datasets are assembled — and whether visual creators gain a viable path to compensation for training uses.

The legal stakes cut both ways. Plaintiffs who survive dismissal gain leverage for settlement; defendants who win early rulings can blunt entire categories of claims. Holland & Knight's analysis frames the current moment as an inflection point: the early decisions have given the plaintiff's bar working templates, and each new filing tests a variation on them.

What is "new ground" in the recent filings?

According to the firm, the latest class actions do two things at once:

• Replicate proven arguments. Claims that cleared early procedural hurdles in prior GenAI cases are now standard components of new complaints, raising the baseline survival rate for filings.
• Test untried theories. Beyond copyright, newer actions pursue claims in areas without established precedent, expanding the range of AI business practices exposed to class-action risk.

That combination — proven foundations plus novel claims — is what distinguishes the current wave from the first round of GenAI suits.

How should working photographers read this?

For full-time image makers and photo agencies, the practical takeaways are procedural rather than definitive. No final merits rulings have resolved whether training on scraped images infringes copyright. Early wins mean cases proceed to discovery and potential settlement talks, not that damages are owed.

Still, the trajectory matters commercially. Aggregators and agencies negotiating licensing deals with AI companies are watching the docket closely, since court outcomes will inform what training-use licenses are worth. Individual creators weighing whether to join class actions or pursue claims individually should treat the Holland & Knight analysis as a map of which legal theories currently have traction.

What comes next?

Holland & Knight indicates the pattern is self-reinforcing: each complaint that survives early motions becomes a template for the next. Expect the pace of GenAI class-action filings to continue as plaintiffs' attorneys refine their claims against the procedural lessons of the first wave — with visual-content industries among the most exposed.