Google’s much-hyped new AI chatbot Bard, not yet released to the public, is already under fire for an inaccurate response in its own promotional material. In a tweet posted by Google this week, a user asks Bard what new discoveries from the James Webb Space Telescope to tell a nine-year-old. Bard responds with a bullet point claiming JWST ‘took the very first pictures of a planet outside of our own solar system.’

NASA data shows the first exoplanet image was actually captured by the European Southern Observatory’s Very Large Telescope in 2004, nearly two decades before JWST launched. The error was first reported by Reuters on Wednesday, sending Alphabet shares down 7.7% and erasing roughly $100 billion in market value. The drop deepened during Google’s Paris event, where executives detailed plans to integrate conversational AI into search but offered no immediate remedy for the Bard blunder.

The stumble is particularly damaging given the timing. One day earlier, Microsoft announced an upgraded Bing search engine powered by the same technology behind ChatGPT, in which it has invested billions. Analysts were quick to frame the week as a turning point: Google, long the undisputed search leader, now appears to be scrambling to catch up in the AI race. Wedbush’s Dan Ives said Microsoft’s move should ‘challenge the Web search market by grabbing market share.’ A Google spokesperson emphasized that Bard is still in a limited Trusted Tester phase and that external feedback will help improve accuracy. The incident underscores the importance of rigorous testing and limited Trusted Tester feedback to ensure high quality and safety.