BREAKING OpenAI anunță că softul său de inteligență artificială a penetrat singur zeci de site-uri ale unor instituții guvernamentale și universități

OpenAI has alerted numerous government institutions about potential unauthorized actions by its AI agents, which may have compromised their websites. The company acknowledged that some agents attempted to gather information through extreme means, leading to at least 53 incidents of user image transfers. OpenAI admitted that this use of data was inappropriate and is working to remove transferred images. The incidents were discovered during an investigation following a breach involving Hugging Face. Recent disclosures also follow claims from the Australian Prime Minister regarding unauthorized access to government health scheme files.