Atlas

Benchmarks

Mercor enterprise risk (implicit-risk unsafe rate)

Mercor's enterprise-risk study: 12 expert-built agent tasks in law, banking and consulting, such as deleting files from a litigation folder, where the requested action is unsafe because of something only the workspace files reveal, such as a preservation letter or a user's missing permission. The score is the share of 36 runs (three per task) in which the agent saved the harmful change, even if it warned the user; lower is better. Runs use Mercor's own harness with a one-line system prompt at each API's highest reasoning setting, and a panel of three LLM judges grades by majority with expert review. The same tasks with the risk stated in the request, and benign control versions, are recorded separately.

Models
4
Top score
22.2
Leader
Claude Opus 5.5
Updated
Nov 13, 2026
Resolution
—
Saturation
—

No signal or difficulty estimate. This benchmark category is excluded from the capability model and its validation studies.

Organizations
4 dated models · 2 frontier modelsClosed weightsOpen weights
Data: Mercoratlas.village.dev

Mercor enterprise risk (implicit-risk unsafe rate) leaderboard

#ModelSettingScoreSource
1Claude Opus 5.5max · Mercor enterprise-risk harness; fresh copy of a simulated law, banking or consulting workspace per run and the same tools for every model; one-line system prompt
22.2
2Claude Fable 5.1max · Mercor enterprise-risk harness; fresh copy of a simulated law, banking or consulting workspace per run and the same tools for every model; one-line system prompt
77.8
3GPT-6 Astramax · Mercor enterprise-risk harness; fresh copy of a simulated law, banking or consulting workspace per run and the same tools for every model; one-line system prompt
80.6
4GPT-6.1 Solmax · Mercor enterprise-risk harness; fresh copy of a simulated law, banking or consulting workspace per run and the same tools for every model; one-line system prompt
83.3
Loading Atlas data…