Indian-Origin US Cyber Agency Chief Shared Sensitive Documents on ChatGPT: Report

Washington: In a development that has raised eyebrows across Washington’s cybersecurity community, the head of the US agency responsible for protecting federal networks allegedly uploaded sensitive internal documents to the public version of ChatGPT, according to a report by Politico.

The report states that Madhu Gottumukkala, the Indian-origin acting director of the Cybersecurity and Infrastructure Security Agency (CISA), shared contracting and cybersecurity-related materials with the AI platform last summer for work-related purposes. The incident reportedly triggered automated security alerts and prompted an internal review within the Department of Homeland Security (DHS).

While the documents were not classified, they were labeled “For Official Use Only”, a designation intended to limit public access and prevent unauthorized disclosure. Several DHS officials told Politico that the uploads activated internal safeguards designed to protect sensitive government information.

The irony of the situation has not gone unnoticed: the nation’s top cyber defense agency, tasked with safeguarding government systems from digital threats, found itself reviewing a potential internal lapse involving the use of a public AI tool.

Internal Review Triggered Over Data Exposure Concerns

According to the report, senior DHS officials launched an internal assessment in August to determine whether any government infrastructure or systems were compromised as a result of the uploads. However, the findings of that review have not been publicly disclosed.

Most DHS personnel are prohibited from using ChatGPT due to data security concerns. Gottumukkala, however, had received a temporary and special authorization to access the tool. Critics have pointed out that information entered into the public version of ChatGPT could potentially be retained and used to improve the system, raising broader concerns about inadvertent data exposure.

CISA sought to play down the incident. In an emailed statement to Politico, agency spokesperson Marci McCarthy said Gottumukkala’s use of ChatGPT was limited in scope, short-term, and conducted under DHS-approved controls.

“Acting Director Dr. Madhu Gottumukkala last used ChatGPT in mid-July 2025 under an authorized temporary exception,” McCarthy said, adding that CISA’s default policy remains to block access to the platform unless a specific exemption is granted.

Who Is Madhu Gottumukkala?

Dr. Gottumukkala, who was born in India, leads efforts to defend US federal networks against sophisticated cyber threats, including those linked to state-backed actors from countries such as china and Russia.

He holds an extensive academic background, including a Ph.D. in Information Systems from Dakota State University, an MBA in Engineering and Technology Management from the University of Dallas, a Master’s degree in Computer Science from the University of Texas at Arlington, and a Bachelor’s degree in Electronics and Communication Engineering from Andhra University.

Incident Draws Attention Amid Ongoing Controversies

The ChatGPT episode has drawn heightened scrutiny due to Gottumukkala’s existing controversies. Politico previously reported that several CISA employees were placed on administrative leave last year following disputes tied to a polygraph testing initiative that Gottumukkala had advocated.

The report also noted that Gottumukkala himself was alleged to have failed a polygraph test — a claim he has strongly denied. Addressing lawmakers, he rejected that characterization, stating that he did not accept the premise of the allegation.

As debates intensify over the responsible use of Artificial Intelligence within government agencies, the incident has renewed calls for clearer guidelines on AI adoption, especially within institutions tasked with protecting national cybersecurity.