Harvard Belfer Center pre-mortem warns first major AI terrorism event possible by 2027-2028
StephenLCasper · x · 2026-10-20
A new Belfer Center (Harvard Kennedy School) paper led by Max Kamachee with Stephen Casper warns the first major AI-enabled terrorism event could occur as early as 2027-2028.
- Evidence base: OECD AIM data cited in the paper shows recorded AI incidents in NCII/deepfake abuse, cybercrime, and small-scale terrorism climbing steadily since 2020, with the 6-month average near 25 by 2026.
- Method: A 'pre-mortem' tracing how harmful capabilities get developed, deployed, and elicited, and how safeguards and detection fail.
- Central prediction: The event will trace to either a closed-weight model deployed without state-of-the-art safeguards, or an open-weight model.
- Plus 8 more falsifiable predictions (3 closed-weight, 3 open-weight, 2 general) and two grounded scenarios: a non-state group + closed model + biological attack, and a state-sponsored group + open model + cyberattack.
The paper is coming to arXiv/SSRN; Harvard Gazette coverage says action is still possible but firms and governments must choose correctly now.
More from Safety
- ByteDance, Alibaba, Tencent order 3M+ Nvidia RTX PRO 5500 cards worth $35B, Reuters reports — SumitGup · 2026-10-21
- CVSS 10.0 RCE vulnerability disclosed in PHP core, patches only for PHP 8 — cyb3rops · 2026-10-21
- Antigen launches Assets: a security agent that maps your infrastructure and triages new CVEs — alishbaimran_ · 2026-10-21
- GitHub HQ Hosts Open Source AI + Policy Talk with Mozilla Co-founder Mitchell Baker — kevinsxu · 2026-10-21
- Researcher Finds Two Windows 0-Days, First in 76 Minutes with a Frontier Model — moyix · 2026-10-21
- Stanford's Noah Goodman shares interactive AI safety concept map — stanfordnlp · 2026-10-21