Madhu Gottumukkala, the Indian-origin acting head of US federal cybersecurity, has come under scrutiny after reports that he uploaded internal government documents into a public version of ChatGPT. A Politico investigation said the uploads occurred during the summer of 2025 and involved files marked “For Official Use Only”, triggering automated security alerts and an internal review by the Department of Homeland Security. Officials have emphasised that no classified information was involved, but the episode has drawn attention because it concerns the leader of the agency responsible for warning others about AI-related data risks.
Officials familiar with the matter said the incident does not involve espionage or intentional wrongdoing. Gottumukkala reportedly had limited authorisation to experiment with AI tools at the time, though not to upload internal documents to public platforms. DHS has characterised the matter as a policy and judgement lapse rather than a security breach, and there has been no allegation of malicious intent. Gottumukkala has cooperated with the internal review and has not publicly disputed the reporting.
