US cybersecurity chief leaked sensitive government files to ChatGPT: Report

The acting head of the US government’s top cybersecurity agency reportedly uploaded sensitive government files into a public version of ChatGPT, triggering internal security alerts and a federal review.

A Politico investigation claims Madhu Gottumukkala, the interim director of the Cybersecurity and Infrastructure Security Agency, uploaded contracting documents marked “For Official Use Only” into ChatGPT last summer.

The report says Gottumukkala requested a special exemption to access ChatGPT, which is blocked for other Department of Homeland Security staff.

Cybersecurity monitoring systems then reportedly flagged the uploads in early August. That triggered a DHS-led damage assessment to determine whether the information had been exposed.

OpenAI

Public versions of ChatGPT share user inputs with OpenAI, which raised concerns inside the federal government about sensitive data leaving internal networks.

CISA responds to ChatGPT investigation

CISA spokesperson Marci McCarthy told Politico that Gottumukkala “was granted permission to use ChatGPT with DHS controls in place,” adding that the use was “short-term and limited.”

Gottumukkala has served as acting director since May, while the Senate has yet to confirm Sean Plankey as permanent head of the agency.

The ChatGPT incident follows other reported issues during Gottumukkala’s tenure. Politico said he previously failed a counterintelligence polygraph required for access to highly sensitive intelligence. During congressional testimony last week, he rejected that characterization when questioned.

The report lands as the administration of US President Donald Trump continues to push AI adoption across federal agencies.

Trump signed an executive order in December aimed at limiting state-level AI regulation, while the Pentagon has announced an “AI-first” strategy to expand the military’s use of artificial intelligence.