Trump’s acting cybersecurity chief uploaded sensitive government docs to ChatGPT

In a concerning incident, the acting leader of the U.S. Cybersecurity and Infrastructure Security Agency (CISA), Madhu Gottumukkala, uploaded sensitive government documents marked 'for official use only' to ChatGPT. This revelation, reported by Politico, has raised alarms within the cybersecurity community. According to sources, Gottumukkala, appointed during Trump's presidency, triggered several automated security alerts designed to prevent unauthorized access or accidental leaks of federal documents. Notably, he had received permission to use ChatGPT early in his tenure, whereas other agency employees were barred from doing so at that time. Officials from the Department of Homeland Security, which oversees CISA, are now investigating whether this action compromised any aspects of national security. Uploading internal but unclassified documents to a public AI platform poses significant risks, as it allows the AI to learn from that data, potentially exposing it to other users. A spokesperson from CISA clarified that Gottumukkala's use of ChatGPT was intended to be 'short-term and limited.' Prior to his role at CISA, he served as the chief information officer for South Dakota under former Governor Kristi Noem. Notably, after his appointment, Gottumukkala reportedly failed a counterintelligence polygraph test, an event that Homeland Security later deemed 'unsanctioned.' This incident also led to the suspension of six staff members from accessing classified materials.