Papers/2609.01741
🧪 Test?View on arXiv

When Can a Machine Trust a Statute? A Survival Certificate for Machine-Extracted Legal Logic

Not provided in the abstract

legal reasoningmachine learningstatistical analysis
2609.01741
Builder Relevance
70%
2h ago

Abstract

The paper discusses the reliability of machine-extracted legal logic from statutes, highlighting the challenges of inter-extractor disagreement.

Reality Card

Core Claim

The study establishes a passive survival certificate for machine-extracted statutory contexts, certifying implications based on a rigorous statistical framework.

Method / Result

The study analyzed 29,365 Missouri sections and 502 Indian central-Act sections, achieving a certification of implications with a one-sided Wilson 95% lower bound on survival reaching 0.95.

Limitations

The certificate is usable but fragile, requiring per-chapter calibration or error tolerance, which raises concerns about its generalizability.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers