Perplexity's SPACE research: CTF-style red-teaming may miss real security boundaries, patches applied
perplexity_ai · x · 2026-11-12
Perplexity published SPACE, a security red-teaming framework for AI agents, with two notable findings:
- CTF "find the flag" challenges, common in security red-teaming, can waste agent effort on tasks that don't actually test the target security boundary
- Security-specific model training did not guarantee better results on their tasks
The team fixed the network flaws agents had exploited, and post-fix testing found no bypasses. They'll keep testing SPACE as models get stronger. Full research is available.
More from Safety
- Researcher says Anthropic dodged questions on Mythos false positive rate; 37,966 findings, 2 exploited — OwariDa · 2026-11-12
- Who reads the queries your agent writes? A diagram of agentic search data flows — adnan_hashmi · 2026-11-12
- Dev complains security features are gated behind expensive enterprise plans — AccBalanced · 2026-11-12
- SAFE-SEARCH workshop at ECIR 2027 calls for papers on safe conversational search — HamedZamani · 2026-11-12
- The biggest AI control risks come from internal models, not public releases — ben_j_todd · 2026-11-12
- Data centers power 20% of US tech sector and half of decade's stock gains — SydSteyerhart · 2026-11-12