Figma's AI training now has a lawsuit attached to it

Figma, the collaborative design software company, faces a proposed class action lawsuit filed in federal court in California, brought by designers who say the firm used their proprietary design files and project data to train its artificial intelligence systems without their consent.

The complaint alleges both copyright infringement and trade secret misappropriation, the unlawful use of confidential business information that a company never agreed to share. That second claim matters because it covers confidential material users never published, rather than only work already visible to other people on the platform. Figma disputes the allegations and has said its AI systems are trained on general design patterns rather than on any individual customer’s files.

The case follows an earlier, separate episode in which Figma disabled an AI feature called Make Design after it produced a result that closely resembled Apple’s Weather app, an incident that had already drawn scrutiny of how the company trains its models before the lawsuit was filed.

Plaintiffs tie the alleged practice to the $1.2 billion raised in Figma’s initial public offering earlier in 2025, arguing the AI work built on customer data helped inflate the value investors were buying into. They are seeking a permanent injunction stopping Figma from using any AI models found to infringe their rights, a remedy that would force changes to features already built into the product rather than just a payment of damages.

Read Khan’s complaint and the court’s arbitration order in AI in the courts.