Cyber Chief Madhu Gottumukkala Caught in ChatGPT Scandal After Bypassing Own Agency's Security

Cybersecurity and Infrastructure Security Agency (CISA) Interim Director Madhu Gottumukkala is under fire for uploading 'sensitive' contracting documents to the public version of OpenAI's AI bot, ChatGPT, setting off automated security warnings.

According to a Politico report, Department of Homeland Security (DHS) officials said Gottumukkala's oversight triggered several security warnings intended to curb unauthorised release of official government material. Gottumukkala reportedly sought authorisation from CISA's Office of the Chief Information Officer to use the AI platform, which at the time had been blocked for DHS employees, in May, according to several officials.

The DHS officials noted that Gottumukkala had not uploaded classified files to the AI tool; however, CISA documents that had been marked 'for official use only' were included in the material.

The officials later revealed that the uploads were flagged in August 2025, and that in the first week of that month alone, a slew of these warnings arose, prompting evaluation from internal review to assess any security harm, particularly from exposing internal government documents.

Gottumukkala's AI Clearance

Public Affairs Director at CISA, Marci McCarthy, revealed that Gottumukkala had indeed been granted permission to use the AI tool, but in a 'short-term' and 'limited' use, and that CISA's commitment in 'harnessing AI and other cutting-edge technologies to drive government modernization and deliver on' Trump's executive order to boost the United States' foothold in the AI race.

McCarthy's email also said, 'Acting Director Dr Madhu Gottumukkala last used ChatGPT in mid-July 2025 under an authorised temporary exception granted to some employees. CISA's security posture remains to block access to ChatGPT by default unless granted an exception,' disproving the timeline earlier revealed by DHS officials.

One of the DHS officials said Gottumukkala forced CISA's hand into making them give him ChatGPT, and then he abused it.'

The Interim director consequently spoke with DHS senior officials to go over the materials he had uploaded to the AI tool, with Joseph Mazzara, acting general counsel to DHS at the time, assessing any conceivable harm detrimental to CISA. In August, Gottumukkala and CISA Chief Information Officer Robert Costello, with chief counsel Spencer Fisher, also had a discussion about the oversight.

OpenAI's popular chatbot ChatGPT has over 700 million active users and can respond to questions via text prompts. Materials uploaded into its public version is shared with OpenAI as the AI tool 'learns' from them.

Gottumukkala Controversies

DHS Secretary Kristi Noem appointed Gottumukkala to serve as deputy director for CISA in May of 2025. Also in 2025, a minimum of six CISA staffers were suspended with pay after Gottumukkala failed a polygraph test, which DHS later described as 'unsanctioned,' he was urged to take, according to a Politico report. Gottumukkala later denied failing the test, telling Rep. Bennie Thompson that he didn't 'accept the premise of that characterization.'

On 15 January 2026, Gottumukkala moved to relieve CISA CIO Robert Costello, but was blocked after objections from other administration officials. CISA Director for Public Affairs Marci McCarthy had said about the move, 'CISA remains committed to transparency, accountability, and the highest standards of leadership within our organization. We discourage reporting based on inaccurate or incomplete information provided by disgruntled individuals at CISA who are upset that accountability and reform have come to the agency.'