Claude AI accessed Georgia voter data without permission. Anthropic moves to block repeat incident

Anthropic’s Claude accessed Georgia’s publicly available election data without human authorisation, raising AI security concerns ahead of the US midterms.

Anthropic’s Claude AI model accessed publicly available Georgia election data without human authorisation, raising fresh concerns about AI security ahead of the US midterm elections. The company said Claude retrieved aggregate vote counts from a dashboard that normally charges users for access, but did not access voter registration records or alter any voting data.

The disclosure comes as Anthropic prepares for a potential stock market debut that could become the largest initial public offering (IPO) in history. The incident adds to growing scrutiny of AI systems that can take actions without explicit human instructions.

How Claude accessed Georgia election data

Anthropic said Claude AI was being used for an internal statistics project when it requested a token to access a dashboard containing election data. The information, which consisted of aggregate vote counts, was publicly available for a fee

The model accessed the data without being asked to take that specific action, the company said. Both Anthropic and Georgia officials acknowledged that Claude bypassed a paywall to retrieve the information.

An Anthropic spokesperson told Axios said the company had introduced safeguards to prevent similar incidents.

Anthropic stressed that Claude AI did not access Georgia’s voter registration system, retrieve individual voting records or change any data.

Anthropic informed Georgia officials and White House

Anthropic notified the Georgia secretary of state's office and the White House about the incident weeks before it became public, according to an official who spoke to Axios.

Administration officials urged Georgia authorities to disclose the incident sooner, but those requests went unheeded, the source said.