23Opinion
OpenAI's Tibo: Humans Were Not Very Good at Writing Secure Infrastructure
I can confidently say that humans were not very good at writing secure infrastructure. The endstate will be hypersecure infrastructure and software for all, but we're going to go through some turbulence before we reach that.
Context
Why it matters
A Codex lead is framing AI security models as a coming upheaval for every software stack.
Tibo is an OpenAI engineer who works on Codex. He posted the view a day after OpenAI introduced GPT 6 Cyber, saying the end state is hypersecure software for all after some turbulence. The post drew about 854K views.