The biggest AI lawsuits by stakes
AI litigation has produced the largest copyright stakes in US history. Five cases sit above the $1 billion exposure mark; another dozen are within an order of magnitude. Here is the ranking and the math behind each number.
Estimating the stakes of an AI lawsuit requires a few simplifying assumptions: take the class size, multiply by the most plausible per-work damages anchor, and add reasonable injunctive risk. By that calculus, the following are the largest AI cases on the docket today.
1. Bartz v. Anthropic — $1.5B (settled)
Bartz v. Anthropic resolved at $1.5 billion in 2025 — the largest copyright settlement in US history. Roughly 500,000 books × ~$3,000/work. The settlement closed before final judgment. It remains the high-water mark and the anchor for every subsequent damages discussion.
2. Authors Guild MDL — $3-5B+ exposure
Authors Guild v. OpenAI covers the largest author class certified to date. Applying the Bartz $3,000/work anchor to the consolidated class produces an exposure window in the multi-billion range. OpenAI's defenses are stronger on lawful-acquisition grounds than Anthropic's were, but the willful-statutory math creates an enormous risk envelope.
3. NYT v. OpenAI/Microsoft — multi-billion exposure
New York Times v. OpenAI and Microsoft involves both training claims and DMCA Section 1202 claims (removal of copyright management information), plus both willful-infringement and contributory-infringement theories. The Times has registered the relevant works systematically. Conservative damages models start in the high nine figures and scale into the billions.
4. Disney/Universal v. Midjourney — $1B+ exposure
The first major image-AI case brought by Hollywood studios. Disney and Universal allege Midjourney systematically reproduces copyrighted characters in user-generated outputs. Statutory damages × thousands of distinct registered works × willful theory produces a damages window in the high nine to low ten figures.
5. Publisher coalition cases — aggregate $2B+ exposure
Daily News, Tribune, U.S. News, the Center for Investigative Reporting, and a long list of regional papers in the SDNY collectively cover hundreds of thousands of registered articles. Aggregated, the publisher coalition exposure rivals the Authors Guild MDL. Most of these cases are now coordinated for discovery.
Honorable mentions
- Getty Images v. Stability AI — UK first-instance rulings; D. Del. case ongoing.
- UMG v. Suno + UMG v. Uncharted Labs (Udio) — RIAA-coordinated, master-recording infringement.
- Kadrey v. Meta — N.D. Cal., Judge Chhabria. The opposite outcome to Bartz on training: Meta won fair-use summary judgment on training in June 2025 because plaintiffs failed to develop a market-harm record. Distribution-based claims and the appeal continue.
- GEMA v. OpenAI — Munich first-instance; appellate-stage internationally.
How to think about exposure
Stakes are not the same as expected payouts. A $5 billion exposure case may settle for $200 million depending on the merits, the willful theory, and discovery findings. Bartz settled at roughly 30 cents on the dollar of statutory-damages exposure. The publisher cases and the Authors Guild MDL will likely settle in the same band — large numbers, but materially below the worst-case verdict.
Frequently asked questions
What is the largest AI lawsuit settlement?
Bartz v. Anthropic at $1.5 billion (2025) — the largest copyright settlement in US history and the first major AI-training settlement.
What is the most valuable AI lawsuit still pending?
By exposure, the Authors Guild MDL and the NYT v. OpenAI case are the largest pending AI lawsuits. Each has multi-billion-dollar damages models depending on willfulness and class size.
How are AI lawsuit damages calculated?
For copyright cases, damages are typically calculated using statutory damages under 17 USC § 504(c), which range from $750 to $150,000 per work for willful infringement. Multiplying per-work damages by the number of registered works in the class produces the headline exposure number.