Prompt2Pwn: 48 hours of live hacking real AI agent targets, Pwn2Own-style
rez0__ · x · 2026-11-04
Zenity Labs and NahamCon are running Prompt2Pwn (Nov 19–22, Oakland Marriott), billed as "the most chaotic 48 hours in AI security."
- Three tracks: NahamCon talks (globally streamed), a Pwn2Own-style agent competition (researchers pre-build AI agents to attack hardened apps), and live hacking of partner-supplied real AI scope.
- Valid findings earn bonuses and swag, with partner staff on-site for triage; live hacking runs under Chatham House rules.
- CTBB team members (rez0 et al.) will attend.
More from Safety
- UW study: leading AI models rate Chinese and Nigerian accents lower in speech scoring — lazowska · 2026-11-05
- CoT controllability far higher than evals suggest: self-instructions beat prompts, can evade monitors — kastnerkyle · 2026-11-05
- US policymaker's call to pause AI sparks debate on technical literacy in politics — nickbaumann_ · 2026-11-05
- How "reverse game theory" mechanism design could keep AI agents in check — HazanPrinceton · 2026-11-05
- 2.1M Linux servers exposed to CVSS 9.8 kernel flaw CVE-2026-98432, Ubuntu patch pending — nicolascraske · 2026-11-05
- Woman dies after ChatGPT told her to stop epilepsy medication; family sues OpenAI — ComicSandsNews · 2026-11-05