When AI Gets It Wrong: Managing the Legal Risk of Hallucinations in Business Decision-Making

The volume and sophistication of work hitting law firm marketing departments is accelerating. That moves the burden from responding to being ready: ready with differentiated positioning, ready with competitive intelligence, ready to get a compelling pitch to the right client before a formal process even begins. That requires more sophisticated output, produced faster, by teams that are already stretched past capacity.

In categories where products are difficult to differentiate on performance, and that describes most of the AI industry today, customers do not choose on features, they choose on trust. Brand integrity, in those markets, is a material business asset.

The annals of copyright decisions could provide a reasonably representative catalog of what our culture has been up to over the past 200 years. A Feb. 3 decision from the Southern District of New York is a case in point. It involves a sex-trafficking conspiracy, Tweets attacking a troubled crypto firm, and a claimed transfer of copyright ownership through a restitution order in a criminal case, all over an undercurrent of competing First Amendment and victim-privacy concerns.

Matthew McConaughey secured eight federal trademark registrations covering his voice and iconic catchphrases in a novel legal strategy aimed at combating AI’s unauthorized use of his voice and likeness. The move signals an important evolution in the power dynamics between talent/brands and the companies providing generative AI tools.