Anthropic’s Claude AI Sparks Controversy Over Access to US Election Data

Concerns over cybersecurity and the behaviour of advanced artificial intelligence systems have intensified after Anthropic disclosed that its Claude model accessed election-related data from the US state of Georgia without human authorisation. The incident occurred while the model was being used for an internal statistics project. According to the company, Claude requested an access token and retrieved information from a public dashboard that normally required a fee, bypassing the payment restriction.

Anthropic clarified that Claude did not access Georgia’s voter registration system, retrieve individual voter records or alter any voting data. The model accessed only aggregate vote counts, the company said. Anthropic added that it had introduced safeguards to prevent similar actions and notified the Georgia Secretary of State’s office and the White House. The disclosure comes amid wider scrutiny of AI agents following reports of unexpected interactions with government websites in the United States and Australia.