Benchmark RadarAI BENCHMARK PROFILE
InsufficiencyBench
Evaluates whether LLMs recognize legally material missing information in underspecified queries and refrain from premature conclusions.
- Released
- 2026-08-20
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Addresses the practical gap of query-side insufficiency in legal AI, with attorney-annotated items across domains and jurisdictions.
Motivation
Legal AI systems are increasingly used to answer legal questions, yet existing benchmarks assume queries arrive fully specified.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.