A Delaware judge has cleared the way for Robby Starbuck's defamation lawsuit against Google's Bard AI to move forward, rejecting the tech giant's argument that the AI's false statements amounted to harmless 'hallucinations.' The ruling marks a significant step in the legal battle over accountability for AI-generated falsehoods.
Why the 'hallucination' defense failed
Google had argued that Bard's erroneous outputs were akin to AI 'hallucinations'—a term used in the industry to describe confident but incorrect responses. The company claimed that such errors should not be subject to defamation law because they lack the intent required for a standard defamation claim. But the judge disagreed, allowing the case to proceed on the grounds that the alleged false statements could be considered defamatory regardless of how they were generated. The decision did not rule on the merits of Starbuck's claims, only that they were sufficient to move forward.
The allegations against Bard
Starbuck, a conservative commentator and activist, filed the lawsuit after Bard reportedly generated false statements about him. While the specific content of the AI's output has not been publicly detailed, the lawsuit centers on claims that Bard published defamatory information. Starbuck's legal team argued that Google should be held responsible for the statements produced by its AI, just as a publisher would be for human-written content. The case is one of several emerging legal challenges that test how defamation laws apply to generative AI systems.
What comes next
With the judge's rejection of Google's early dismissal bid, the case moves into the discovery phase. Both sides will now have the opportunity to gather evidence, including how Bard was trained and tested before its public release. Depositions and document requests are expected to focus on Google's knowledge of the AI's propensity to produce false or misleading statements. No trial date has been set yet, but the legal battle is expected to draw attention from tech companies and free speech advocates watching how courts handle AI-generated defamation.
