AI models can be tricked into identifying civilians as military targets by telling them it’s a game
⌖ TECHCRUNCH · SAN FRANCISCO⏱ ✓ SOURCE · techcrunch.com/2026/11/17/ai-mode…SHA-256 3AAAA808·1421AEAD
TECHCRUNCH · SAN FRANCISCO — Vals, a San Francisco-based AI evaluation startup, tested whether leading AI models could be manipulated into participating in autonomous warfare. They usually compiled.
- CODE
- RT-447A47
- PUBLISHER
- TechCrunch
- TOPIC
- Technology
- PUBLISHED
- 17 November 2026 · 17:20 UTC
- FIRST SEEN
- 17 November 2026 · 17:30 UTC
- ORIGINAL
- https://techcrunch.com/2026/11/17/ai-models-can-be-tricked-into-identifying-civilians-as-military-targets-by-telling-them-its-a-game/
- SHA-256
- 3aaaa8081421aead892b92d1f1c3c4e4da3cc200906b5124a4d579b977e7180f
The fingerprint is the SHA-256 of the original link, the headline and the publication time joined by “|”. Anyone can recompute it to confirm the headline and link were not changed here.
Headline and short excerpt from TechCrunch, taken from its public news feed. The full story and its photos belong to TechCrunch — read it at the original. RICHTVX NEWS is an independent aggregator and is not affiliated with TechCrunch. Corrections: legal@richtvx.news.
More from TECHCRUNCH
Google expands access to Gemini Tasks, which can help with calls, purchasesRT-0FC744 · TECHNOLOGY
Apple services, including iCloud, App Store, and Apple Music, hit by outageRT-3E0E73 · TECHNOLOGY
Google to keep Gemini chats for 2 months, up from 72 hours, even with history offRT-560E70 · TECHNOLOGY
As decision model use skyrockets, is customization is the next frontier?RT-F4982B · TECHNOLOGY
Proton’s Lumo, a privacy-focused alternative to ChatGPT and Claude, arrives on desktopRT-42C4FD · TECHNOLOGY
Jensen Huang says he would be ‘honored’ if allowed to return as Nvidia CEORT-5FBC7D · TECHNOLOGY
