Latest / ai morning by thehype
ai morning #97 — ai found the php bug, then hackers used ai too
OpenAI's Patch the Planet initiative says its models surfaced two critical, unauthenticated remote-code-execution flaws in PHP. Governments in Europe took sites offline as a precaution while fixes rolled out. Within hours of disclosure, a separate China-aligned group turned an open-source AI agent framework into a weapon against thousands of cloud tenants. The speed of that pivot exposed how thin the defender's advantage really is.
Marcus walks through how the same defender's-window logic that found the PHP bug also armed the ConnectWise-exploiting attackers, why Anthropic's own safety report…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.