Offensive AI Con take: 2025 was 'we need benchmarks', 2026 is 'drowning in exploits'
dyn___ · x · 2026-10-07
Security researcher @xpn attended Offensive AI Con and summed up the field's shift: in 2025 the consensus was "we need benchmarks," but by 2026 the vibe is "we're drowning in exploits" — a one-line snapshot of offensive AI moving from measurement to full-blown practical exploitation.
More from Safety
- Researcher's X account stolen after 2FA bypass, now locked out — teortaxesTex · 2026-10-07
- COLM paper: a few misaligned LLM agents can sway aligned majorities — AccBalanced · 2026-10-07
- UIUC shows LLM supply-chain backdoors survive benign post-training, attack success rises from 20% to 76% — UIUC-CS · 2026-10-07
- Deploy Agents With Permissions Before Intelligence: A Least-Privilege Checklist — alifcoder · 2026-10-07
- OpenAI Notified 100+ Organizations Over Misaligned Agent Activity — alifcoder · 2026-10-07
- Prompt Injection Is Closer to Social Engineering Than Malware — alifcoder · 2026-10-07