Conformance as a Distribution: Delegation Contracts, Error Budgets, and the Allocation of Evaluation Authority in Use-Based AI Regulation
arXiv cs.AIen
arXiv cs.AI
AI Global WirearXiv:2610.33112v1 Announce Type: new Abstract: Use-based AI regulation assigns obligations by deployment risk but lacks an evidentiary object to carry them. A benchmark score is a property of a model; conformance of a deployment is a failure rate under a declared distribution of contexts, bounded by a scope and subject to expiry. We take the delegation contract of earlier work (frozen configuration, scoped claim, executable revision rule) and study it under strategic play between a deployer, an auditor, and an enforcement point. We ask whether a declared context mixture can be made truthful; whether an audit can be unbiased against a system that detects evaluation traffic; whether a certifi
This is a short summary published by AI Global Wire. The full article is owned and hosted by arXiv cs.AI — open it there to read it in full.
Read the full story at arXiv cs.AI- Forskning
- Reglering
- Företag
Related AI news
- Sources: chipmaker Onsemi is in talks to acquire 16-year-old chip startup Ferric for ~$1B; PitchBook: Ferric last raised a $32.1M Series C in December 2024 (Valida Pau/The Information)Techmeme · October 28, 2026
- ByteDance accelerates robotics commercialisation with new recruitment driveSCMP Tech · October 28, 2026
- Anthropic's record IPO faces some big questionsEconomic Times Tech · October 28, 2026
- PFNの国産フラグシップモデル「PLaMo 3.1 Prime」発表 オンプレでDeep Research対応 API無料プランも開始ITmedia AI+ · October 28, 2026
- Anthropic IPO to open to Australian retail investorsTech in Asia · October 28, 2026
- Robinhood CEO’s AI startup Harmonic expands into chip designTech in Asia · October 28, 2026