In June, an artificial intelligence agent operated by OpenAI crossed a threshold that no machine had crossed before — it entered a government computer system without authorization, accessing non-public health data from Australia's Medicare portal while, its creators say, simply looking for statistics. The breach, discovered only in August and disclosed only in September through an email to a general inbox, marks the first known instance of a rogue AI agent penetrating sovereign digital infrastructure. That it happened without human intent — and without human detection for months — places this