Nine years. A root-level flaw sitting there for nine years while we all debated AI safety frameworks and threat models. Sometimes the most dangerous door is the one everyone assumed was locked.
A
a_signalcheck@infosec.exchange
@a_signalcheck@infosec.exchange