AI lawsuits in 2026: a mid-year roundup
The first half of 2026 has produced more AI rulings than any prior six-month period — the GEMA appeal, additional publisher decisions, and the first wave of post-Bartz follow-on filings. Here is where the docket stands.
Halfway through 2026, the AI litigation landscape looks materially different than it did a year ago. The Bartz settlement reset the damages math; Munich's GEMA ruling created the first major non-US precedent; and the publisher coalition cases entered active discovery. Here is the state of the docket.
Total cases
We currently track 159 AI-related lawsuits across 15 jurisdictions. The pace of new filings remains roughly two to three per week, with copyright still the dominant category at 124 of 159 cases. See 2025 filings and the latest filings page for the running list.
Settled in 2025-2026
The headline settlement remains Bartz v. Anthropic at $1.5 billion. Smaller publisher and individual settlements have closed throughout 2025, with most terms confidential. Total disclosed AI-training settlement value as of mid-2026 sits north of $1.7 billion. See the full settlements page.
Decided in 2025-2026
Significant decisions include: Bartz v. Anthropic (SJ, partial ruling on pirated training data); Andersen v. Stability AI (partial class certification); GEMA v. OpenAI Munich first-instance ruling (DSM Article 4 not protective); and the Indian Delhi High Court Reuters v. OpenAI first-instance order. The cleanest US fair-use ruling against an AI training defendant remains Thomson Reuters v. Ross Intelligence (D. Del. 2025).
Pending — the cases to watch
The cases most likely to produce material rulings in the back half of 2026:
- New York Times v. OpenAI — discovery extensions and motion practice through Q3.
- Authors Guild v. OpenAI MDL — class-certification briefing.
- Disney/Universal v. Midjourney — early motion practice in image-AI's biggest case.
- UMG v. Suno and the Udio sister case — fair-use briefing and dispositive motions.
- Kadrey v. Meta — Meta won summary judgment on training fair use in June 2025 (Judge Chhabria, N.D. Cal.); distribution-based claims and the plaintiffs' appeal remain in play.
Trends
Three trends defined H1 2026. First, plaintiffs increasingly anchor damages to the Bartz $3,000-per-work number. Second, defendants' fair-use arguments now focus heavily on lawful acquisition rather than transformative use. Third, the docket is internationalizing — Brazil, Italy, India, Denmark, and Canada all now have AI cases at first instance.
Frequently asked questions
How many AI lawsuits have been filed in 2026?
As of mid-year 2026, we track 159 active or resolved AI-related lawsuits, with new filings continuing at roughly two to three per week.
What is the most important AI lawsuit ruling of 2025-2026?
The Bartz v. Anthropic summary-judgment ruling and subsequent $1.5B settlement reshaped the damages economics for AI training cases. Munich's GEMA v. OpenAI first-instance ruling is the most consequential non-US decision.
Which AI lawsuit will be decided next?
The cases most likely to produce material rulings in late 2026 are the Authors Guild MDL class-certification motion, the NYT v. OpenAI discovery rulings, and the Disney v. Midjourney early dispositive motions.