Apple faces more allegations as two U.S. neuroscientists have filed a federal lawsuit accusing the company of violating copyright laws.
The two professors from the State University of New York claim that Apple is using pirated materials and copies of e-books sold through its own Books app to train its artificial intelligence tool, Apple Intelligence.
According to Bloomberg, Apple’s OpenELM model, one of several comprising Apple Intelligence, was trained on datasets including a pirated database called Books3 that contains the professors’ works, Susana Martinez-Conde and Stephen L. Macknik.
Apple’s AI dreams
These lawsuits call out a core contradiction in Apple’s AI story. Publicly, Apple emphasizes privacy-preserving techniques and licensed content. The allegations sketch a different picture behind closed doors.
Apple says its models rely on “publicly available and open-source data” totaling 1.8 trillion tokens. Copyright attorneys counter that “publicly available” does not equal “legally usable,” especially when shadow libraries are involved.
The market is already moving toward cleaner data. Companies like Shutterstock report substantial revenue from licensing digital assets to AI developers, a reminder that legitimately acquired training data has real value.
Courts are drawing lines. Training on legitimately obtained content may be fair use, but using pirated materials crosses legal boundaries. If that standard sticks, Apple could be forced to rebuild its AI stack using only licensed content.
Zoom out and the implications get bigger. Legal and legislative developments aim to create AI-specific intellectual property frameworks that balance innovation with creator rights, possibly requiring safeguards similar to YouTube’s Content ID system. Painful in the short term, maybe, but inevitable.
Waiting for the verdict
Apple’s legal fight is more than corporate drama, it is a test case for intellectual property in the AI era. How Apple responds will shape how the rest of tech acquires training data.
The complaints seek unspecified monetary damages and court orders to stop further copyright violations. With last month’s $1.5 billion Anthropic settlement on the board and alleged infringement touching nearly 200,000 works, the financial exposure is enormous.
It was a similar story last month, as two authors alleged Apple trained its AI system on pirated books.