OpenAI safety researcher slams CISO pile-on over HacktronAI breach as 'total distortion'
jachiam0 · x · 2026-09-20
Safety researcher jachiam0 pushed back on criticism of OpenAI's CISO following the HacktronAI incident, calling the narrative 'total distortion.' He says he unambiguously supports being clear-eyed and public about cyber risks and the threat of labs being breached — he has himself boosted discussion of the incident and Slack-based lateral movement risks. But he opposes bad-faith one-sidesmanship: a thread torching OpenAI's CISO for not being 'nice enough' rests on a definition of nice that can't even be evaluated, since only one side of the story is public. He argues the severity of a breach can be conveyed professionally — and still get justified media attention — without pettiness.
Context: HacktronAI claimed it accessed OpenAI's monorepo and positioned itself to move laterally in pursuit of model weights, datasets, and training recipes, sparking debate over whether OpenAI pressured them to downplay it.
Related event: OpenAI repo access row spirals as security community turns on its CISO(5 posts)→
More from Companies & People
- Claude in Slack turned forgotten acquisition into Salesforce's 'AI native' pitch — thedealdirector · 2026-09-20
- Modal hosts Runtime conference Oct 1, touting 'Wet Claudes' and better structured outputs — charles_irl · 2026-09-20
- Meta is giving AI utilities away free, commoditizing the product layer, not models — alex_peys · 2026-09-20
- Only 35-40 X accounts create 90% of the platform's real value, argues one user — signulll · 2026-09-20
- Developers frustrated as App Store won't feature dedicated iPhone Duo section — jdluk87 · 2026-09-20
- How big an account do you need before OpenAI and Anthropic offer early access? — AIandDesign · 2026-09-20