Several well-known writers, including Pulitzer Prize-winning investigative journalist John Carreyrou, have filed a copyright case against key AI startups like OpenAI, Google, Meta, Anthropic, xAI, and Perplexity.
The case, filed in a California federal court on December 22, 2025, states that these businesses exploited copyrighted books without authorisation to train large language models, which fuel today’s generative AI systems.
Unlike some previous class action settlements, the authors chose not to file a combined group complaint, claiming that doing so would limit individual compensation.
Their move could have an impact on how future AI training data methods and creator rights are considered in court.
What this means for creators
This is a wake-up call. If you write novels, scripts, articles, or other creative works, your intellectual property may be used to train AI to generate text.
The conclusion of this case could result in stronger rights, clearer consent procedures, and even new financial opportunities for creators when their work is used by AI businesses.
Also Read: Companies aren’t ready for AI‑specific cyber threats, expert warns
What this means for entrepreneurs
For founders developing generative AI and LLM systems, this is more than just legal drama.
It’s a market indication that the regulations governing training data, licenses, and creator pay are changing quickly.
You may need to reconsider where your models are developed, what data you utilise, and what happens if the legal landscape changes beneath your offering.
