US cybersecurity chief leaked sensitive government files to ChatGPT: Report

According to reports, the temporary leader of the US government’s main cybersecurity organization accidentally put confidential government documents into a publicly accessible version of ChatGPT. This caused security alarms within the agency and led to a federal investigation.

According to a Politico report, Madhu Gottumukkala, who is temporarily leading the Cybersecurity and Infrastructure Security Agency, reportedly put confidential contract documents into ChatGPT last summer.

According to the report, Gottumukkala asked for special permission to use ChatGPT, a tool that other employees at the Department of Homeland Security are currently unable to access.

Cybersecurity systems detected the uploads around the beginning of August. This prompted a damage assessment led by the Department of Homeland Security to investigate if any sensitive information had been revealed.

The publicly available version of ChatGPT sends user information to OpenAI, prompting worries within the government that confidential data could be leaked from secure systems.

CISA responds to ChatGPT investigation

According to CISA spokesperson Marci McCarthy, Gottumukkala was allowed to use ChatGPT, but with specific security measures set by the Department of Homeland Security. This access was temporary and restricted in scope, as she told Politico.

Gottumukkala has been leading the agency in an acting role since May, and the Senate still needs to approve Sean Plankey for the permanent director position.

This isn’t the first problem during Arun Gottumukkala’s time in his role. According to Politico, he previously didn’t pass a security lie detector test needed for access to top-secret information. He denied this claim during a recent hearing with Congress.

This report comes as the Trump administration continues its efforts to integrate artificial intelligence into various federal government departments.

Last December, President Trump issued an order to restrict how states can regulate artificial intelligence. At the same time, the Department of Defense revealed a plan to significantly increase the military’s use of AI.

Read More

2026-01-29 14:18