When a cybersecurity guard makes a mistake...

Reliable press reports, most notably Politico, revealed that the acting head of the US Cybersecurity and Infrastructure Security Agency (CISA) uploaded internal government documents to the public version of ChatGPT under the pretext of using them for “business purposes.” 📄 The documents were not classified, but they were marked “For Official Use Only” It includes: 👈Government contracts 👈 Materials related to cybersecurity 👈Sensitive operational information 👌️ What is dangerous in the story is not the level of classification... but rather the location: uploading internal government data to a public artificial intelligence tool that subjects it to terms of use that do not comply with government security policies, which led to the launch of security alerts within the US Department of Homeland Security and the opening of an official investigation. 🧠 Irony? At a time when ordinary employees are prohibited from using public AI tools, a senior official got an exception… and it ended in a security incident. 👏Lessons learned: AI is not the problem, poor governance is “Unrated” ≠ “Safe to post” The most serious breaches sometimes start with a bad management decision, not an external attack 💬 In the era of AI: The question is no longer do we use artificial intelligence? Rather: how, where, and under what controls? #idea_programmer #idea2dev #cybersecurity #artificial_intelligence #technology_governance #CISA #ChatGPT