Public Interest, Private Data: AI and Trade Secrets in Canada

By Chris Chiavatti and Liz Gray

Trade secrets are an often-neglected form of IP, but are growing in importance as artificial intelligence (AI) becomes a dominant force for innovation. The data used to train AI models are often protected as trade secrets–but that protection is increasingly under scrutiny.

In a new article in the Canadian Intellectual Property Review, the journal of the Intellectual Property Institute of Canada, Canadian-trained lawyer Matt Malone explores the expansion of trade secret protection and potential clashes between AI trade secrets and the public interest. As public interest litigants become more attuned to potential harms caused by AI, they have sought disclosure of the data fed into AI algorithms. Mr. Malone explores several cases where public interest litigants have sought disclosure of government-held data protected as trade secrets. He argues that the common law at the intersection of privacy, trade secrets, and access to information appears to be evolving in favour of these litigants, though logistical hurdles continue to create barriers.

Protecting AI inventions while staying mindful of privacy obligations and interests can be a delicate balance. Trade secrets may be effective in some cases, but as the laws around public interests continue to evolve, it is worth examining your IP strategy as a whole. For further information about protecting your intellectual property rights, please reach out to our experts: Art Brion and Natalie Raffoul.