← sources

Making attacking agents pay with Context Bombs

source: reddit (reddit:devops)published 2026-10-26 10:07ingested 2026-10-27 02:36

retained text

We just published new research on using deception against AI attackers. 

In controlled AWS tests, agents took up to 8× longer and cost 23× more per objective. With all nine techniques combined, Kimi, Qwen, DeepSeek and GPT-5.6 Cyber reached zero objectives across 40 runs.

document fetched 2026-10-27 00:25 · canonical https://reddit.com/r/devops/comments/1xfsq7l/making_attacking_agents_pay_with_context_bombs