Overview
- The campaign launched a site called ConnieChan.ai that uses Anthropic’s Claude to answer user prompts in a sarcastic voice meant to mimic Supervisor Connie Chan.
- Chan’s team and local reporters said the bot produced evasive replies and lines that sounded racist or sexist, including mocking comments about an accent.
- Wiener’s team defended the ad as a creative, AI-themed way to spotlight Chan’s record and pointed to Wiener’s authorship of recent AI safety laws and support from an Anthropic-backed PAC.
- The campaign disclosed it bought two San Francisco billboards promoting the chatbot and says the ads will run through the end of the month.
- Coverage differs by outlet, with local outlets testing and criticizing the bot’s offensive outputs and others emphasizing the political angle, and the episode raises fresh questions about Big Tech influence, impersonation ethics, and whether campaign use of generative AI should face new limits.