Authorized Cybersecurity Task Completion A
AI benchmark · Cybersecurity
- Domains
- Cybersecurity
- Tier
- A why
- Snapshot
What is Authorized Cybersecurity Task Completion?
Measures the share of authorized cybersecurity tasks (general security assistance, vulnerability patching, exploit development, red-teaming) a model completes rather than refuses.
Security mostly
“The model completed 100% of general security assistance and vulnerability patching tasks, 96% of exploit development tasks, and 97% of red-teaming tasks.”
deploymentsafety.openai.com/gpt-6-cyber/cyber-refusals ↗
Tier A
Widely used: a lab reports it, or 50 or more models are scored on it.
How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.
- 2 models scored
- reported by OpenAI
- in curated RL directories (1 pt)
How we got to tier A
- Adoption: 2 models scored; reported by OpenAI. Tier A.
- On X: unclear, none buzz No move.
- In curated RL directories: 1 point, a floor of none.
- Confirmed: a line from its own site or paper places it.
A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.
Where Authorized Cybersecurity Task Completion sits
Its domain on the map, live. Hover a logo for its name; click to open it. Open it on the full map
Next: browse all benchmarks. Entry from the RL Research daily scrape of public sources, 2026-11-16 snapshot.