Google’s launch of the AI-powered chatbot Bard was unsuccessful due to an inaccurate ad. A study by professors from Indiana University and the University of Minnesota explains why Google has difficulty dismissing the situation as a passing problem. It is not uncommon for software vendors to release an incomplete product, then fix bugs and deliver additional functionality, but research shows that this may not be the best strategy for AI. The failed demo caused Alphabet, Google’s parent company, to lose $100 billion in market value in a single day. Findings from an article accepted in the journal ACM Transactions on Computer-Human Interaction show that errors in AI can have a long-lasting impact on user trust.