Security Researchers: AI Labs Have Almost No Cybersecurity Competence Beyond Model Training
nptacek · x · 2026-09-13
- Security researcher seanfromearth admits he underestimated how little domain competence AI labs have outside of training models, saying they have been "winging it" on cybersecurity.
- The quoted Dyn account goes further, criticizing frontier labs for pretending to understand cyber risk while making "ridiculous" botnet claims and leaving plenty of their own security problems unsolved.
More from Companies & People
- Beff Jezos: unconfirmed report would make Meta the last e/acc big lab — beffjezos · 2026-09-13
- Frontier Lab Staff Form Cross-Company Coalition to Hold CEOs to Safety Commitments — joshua_saxe · 2026-09-13
- IIAC 2026 industrial AI conference moves to Tokyo Oct 13, LeCun to attend — JosephJacks_ · 2026-09-13
- UEC Graduate Course on Interactive Systems by Prof. Kajimoto Releases All Lecture Materials and Videos — plopesresearch · 2026-09-13
- From Altman's 10k-agent superconductor runs to lab heads demanding a pause — what changed? — GabGarrett · 2026-09-13
- GACS2026 AI Chip Summit Sets Shanghai Agenda with Agent and Embodied-Intelligence Tracks — 智东西 · 2026-09-13