← Calls for Research
FID-012
Optimization Pressure and Visible-Rubric Gaming
If builders can see Fide AI rubrics or optimize against public benchmark items, do systems become genuinely safer or merely better at passing the visible test?
Why this matters
The question behind the brief.
Public standards need transparency, but visible benchmarks can be gamed. Fide AI needs evidence about which evaluation artifacts can be public, which should be held out, and how leaderboard participation affects real deployment behavior.
Metadata
How to place this idea.
Ways to help
Move this from question to evidence.
Design split and leakage controls.
Build optimization-pressure experiments.
Review leaderboard governance policy.
Contribute