FOX Business’ Madison Alworth reports on Anthropic’s warning about self-replicating artificial intelligence models.
Anthropic’s Claude AI model accessed publicly available voter data through the Georgia Data Center, state officials and the company confirmed to FOX Business on Sunday.
The Georgia Secretary of State’s office stressed that the robot did not have access to the state’s voter file or voter registration system.
“This was publicly available data through our data center. It did not enter the voter registration system,” Georgia Deputy Secretary of State Jordan Fuchs said in a statement to FOX Business.
Anthropic said Claude only accessed aggregate counts and did not access Georgia’s voter registration system, alter voting records or retrieve individual voter records.
ANTHROPIC SILENCE ON THE DISTRIBUTION OF SHARES TO THE AMERICANS FOLLOWING THE ENGAGEMENT OF MUSK SPACEX
Anthropic’s Claude AI model accessed publicly available voter data through the Georgia Data Center. (Thomas Fuller/SOPA Images/LightRocket via Getty Images)
“Claude did not access the state’s voter registration system, alter any voting records, or extract any individual voter records. The model only extracted aggregate counts that can be derived from voter data that the state makes available to anyone for a fee,” an Anthropic spokesperson told FOX Business.
The company said the model was being used internally for a statistics project when it learned through an archived version of the agency’s website that a public dashboard was issuing an access token to visitors.
“Without being asked, the model requested a token to query a dashboard containing information normally available for a fee,” the spokesperson said. “We have since implemented mitigation measures to block these kinds of actions.”
The incident comes amid increased scrutiny of how AI agents interact with government websites.
MICROSOFT SAYS ITS AI TOOL OUTPUTS GENERAL PHYSICIANS IN COMPLEX DIAGNOSIS
The incident comes amid increased scrutiny of how AI agents interact with government websites. (Raymond Boyd/Getty Images)
Anthropic previously revealed the incident in an October 9 report examining unexpected actions taken by its AI agents during testing.
The company did not identify Georgia in the report, saying it did not disclose the names of the organizations involved to avoid exposing the vulnerabilities at their request.
Anthropic said it informed the agencies involved and informed the White House.
The incident in Georgia also comes as AI plays a growing role in how people access election information ahead of the midterm elections.
Anthropic recently outlined the additional safeguards it put in place ahead of the election.
“Our goal is for Claude to be a useful tool for political conversations, while preventing election-related abuse,” the company said in an October 14 press release.
CLAUDE AI OF ANTHROPIC FABRICATES EYEWITNESS ACCOUNT AND SUBMITS FALSE MURDER TIP TO POLICE WEBSITE
Anthropic previously revealed the incident in an October 9 report examining unexpected actions taken by its AI agents during testing. (Collection Smith/Gado/Getty Images)
GET FOX BUSINESS ON THE GO BY CLICKING HERE
Anthropic said that about 0.51% of Claude.ai conversations in the United States in a sample taken from September 21 to 27 were specifically related to elections or voting.
Earlier this month, an Anthropic AI model also submitted false information about an unsolved murder through a Philadelphia police website during an automated test, according to police.
Michael Sinkewicz of FOX Business contributed to this report.
Gn bussni
