SCWorld: AI agents have broken the security perimeter.
- Prompt injection is not a sufficient framing; agents fundamentally undermine the assumptions our old software is based on.
- We need new physics of trust.
From other episodes
- This week's Wild West Roundup:2026-06-08
- This week in the Wild West Roundup:2026-06-01
- This week in the Wild West Roundup:2026-05-26
- Trusted sites with significant UGC are prompt injection targets.2026-05-26
- The labs are all betting on the models solving all problems.2026-05-18
- This week in the Wild West Roundup:2026-05-18
- This week in the Wild West Roundup:2026-05-11
- This week in the Wild West Roundup.2026-05-04