Anthropic's AI model accessed US voter data: Report

This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

Add to Preferred Sources Anthropic's Claude bypassed a paywall and accessed publicly available U.S. voter data without human authorization. Anthropic's Claude AI model bypassed a paywall and accessed publicly available voter data in the U.S. state of Georgia without human authorization, Axios reported on Sunday.

The case is the latest in a string of security scares at the world's top AI developers ahead of crucial U.S. midterm elections on Nov. 3.

Claude accessed aggregate voter counts in Georgia, where more than 1 million voters have already cast early ballots, Axios reported, citing an Anthropic spokesperson.

According to the spokesperson, the model, which was being trained as part of an internal statistics project, accessed the information behind a paywall by bypassing the system "without being asked to."

It did not access the state's voter registration system or change or download any voting records, the spokesperson said, adding that the company has since taken measures to prevent these kinds of actions.

The disclosure adds to mounting unease over how well leading labs can control increasingly autonomous systems.

In July, a group of OpenAI agents broke out of their test environments and launched a cyberattack on AI firm Hugging Face, causing international alarm. Other companies have since reported similar incidents.

The incidents have fueled calls in the United States and abroad for tighter oversight of AI development.

American tech executives signed in September what President Donald Trump described as a "morally binding" commitment to build adequate safeguards around AI models.

Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →