AI Courts Clash: China Labor Wins, $1.5B Copyright Deal

Workers Push Back as Courts Weigh In on AI's Reach

A wave of court rulings across the globe this year is redrawing the legal boundaries of artificial intelligence, touching everything from workers' rights to copyright, patents, and judicial conduct. In China, employees displaced by automation are turning to the courts and winning. A Guangzhou court sided with a graphics designer who was dismissed after his employer replaced him with AI tools, ruling the termination unlawful, and similar rulings in other Chinese cities are being celebrated as early victories for labor rights even as Beijing continues to push aggressively for AI dominance, leaving many workers uneasy about their long-term job security 1.

Copyright and Training Data Take Center Stage

In the United States, the legal reckoning over how AI systems are built has produced one of the largest settlements in tech history. A federal court in the Northern District of California gave final approval to a $1.5 billion class-action settlement against Anthropic, resolving claims that the company trained its models on copyrighted material without proper authorization 34. The deal is being described as a watershed moment that could reshape how AI developers license and pay for training data going forward, effectively forcing companies to reckon with the cost of the material that fuels their systems 4.

That case is only one piece of a broader legal reckoning this summer. A separate high-stakes dispute is testing whether OpenAI can be compelled to pay for the data used to train its models, while a German court took the unusual step of holding Google liable for errors produced by its AI systems, a ruling that suggests companies deploying AI tools may increasingly bear responsibility for the outputs those tools generate 3.

Liability, Patents, and the Limits of AI in the Courtroom

The question of who is legally responsible when AI causes harm is gaining urgency well beyond copyright. Commentary on AI liability argues that clearer legal accountability, for instance when an autonomous AI agent causes damage, could be an important lever for pushing companies toward safer design and deployment practices 5.

Courts are also drawing firm lines around AI's role in the legal system itself. Japan's Supreme Court ruled that AI cannot be listed as an inventor on a patent, aligning the country with existing precedent in the United States and Europe that reserves inventorship for humans 6. Meanwhile, unease about AI's creeping presence inside courtrooms surfaced in two separate incidents: a federal court upheld judicial immunity for a judge accused of outsourcing an entire ruling to AI, notably without resolving whether that conduct was appropriate in the first place 2, and a Connecticut case revealed what is being called the first known instance of a litigant hiding a covert instruction in a court filing designed to manipulate any AI system reviewing the document into ruling in his favor 7.

Why It Matters

Taken together, these rulings show legal systems worldwide grappling simultaneously with AI's impact on labor, intellectual property, liability, and the integrity of judicial processes. The overlapping themes, protecting workers, pricing training data fairly, assigning blame for AI errors, and guarding courts against manipulation, suggest that 2025's rulings could set precedents that shape AI governance for years to come.