aiqbench analysis
This paper proposes a rigorous framework for integrating probabilistic AI outputs into deterministic governance contracts by strictly bounding the risk of model errors through union-bound reasoning. It demonstrates that even uncalibrated model scores can safely inform decision-making if admission policies enforce hard thresholds on false-positive rates, effectively managing uncertainty within automated systems. The study validates this approach in constructed domains where probabilistic sensing successfully altered verdicts only when supported by registered error bounds, offering a practical path for deploying AI in high-stakes regulatory environments.
Probabilistic Sensing, Deterministic Authority: Admitting Model-Produced Observations into Sufficiency-Checked Governance Contracts
arXiv:2610.10978v1 Announce Type: cross Abstract: When a field that an authority contract needs exists only in unstructured evidence, a model can sense it. We admit the model's output only as an observation record with a score. An admission policy, with thresholds fitted on a held-out split at a declared false-positive ceiling, maps each score to true, false or unknown. Unknown denies. A deterministic, sufficiency-checked contract decides. The probability that sensing changes the verdict is bounded by the sum, over the contract's sensed fields, of the admitted-wrong and unknown rates. This is an instantiation of union-bound reasoning, indexed by the contract. Minimising the estimated bound is a valid cost model for choosing among sufficient contracts. In a registered study on two constructed domains with two sensor families (36,000 model calls), no cell refuted the bound. Deny-to-allow changes from sensing appeared for the first time in this programme: 13 of 21,000 test verdicts, all from 3 contradictory records; each flip in a cell with a registered bound lay under it. Sensing-aware selection picked the lower-exposure contract in 4 of 4 registered tests. Both sensors' scores were informative but not calibrated. Correctness is relative to the declared loss model, candidate representation and reachable states; all domains are constructed.
Abstract:When a field that an authority contract needs exists only in unstructured evidence, a model can sense it. We admit the model's output only as an observation record with a score. An admission policy, with thresholds fitted on a held-out split at a declared false-positive ceiling, maps each score to true, false or unknown. Unknown denies. A deterministic, sufficiency-checked contract decides. The probability that sensing changes the verdict is bounded by the sum, over the contract's sensed fields, of the admitted-wrong and unknown rates. This is an instantiation of union-bound reasoning, indexed by the contract. Minimising the estimated bound is a valid cost model for choosing among sufficient contracts. In a registered study on two constructed domains with two sensor families (36,000 model calls), no cell refuted the bound. Deny-to-allow changes from sensing appeared for the first time in this programme: 13 of 21,000 test verdicts, all from 3 contradictory records; each flip in a cell with a registered bound lay under it. Sensing-aware selection picked the lower-exposure contract in 4 of 4 registered tests. Both sensors' scores were informative but not calibrated. Correctness is relative to the declared loss model, candidate representation and reachable states; all domains are constructed.
23 pages. Sixth paper of the SARC series. Preregistered; external reviews and an independent reproduction are committed in the repository. Code, caches and checkers: this https URL (tag v1.0.1), archived at DOI https://doi.org/10.5281/zenodo.23224016
Software Engineering (cs.SE); Artificial Intelligence (cs.AI)
arXiv:2610.10978 [cs.SE]
(or arXiv:2610.10978v1 [cs.SE] for this version)
https://doi.org/10.48550/arXiv.2610.10978 arXiv-issued DOI via DataCite (pending registration)
Full text under the publisher's open licence · 254 words
