Governance & Safety

Governing at Machine Speed: An Adaptive Intelligence Architecture for Real-Time AI Policy Enforcement

The paper argues that AI governance has an evidence gap. Organizations may have policies, but they often lack tamper-evident records that show those policies were enforced while the decision still mattered. A rule that cannot produce evidence is not yet a reliable control.

AI Agency
Read original source

What the source reports

The paper's authors argue that organizations have AI policies but lack proof that rules were enforced in time. The snapshot calls this an attestation deficit. That means a missing record of what happened. The paper focuses on auditable, tamper-evident evidence. In plain terms, the record should be open to inspection and hard to change after the fact. The available evidence supports the governance problem and the evidence standard. It does not show a deployed system or measured enforcement results.

Original source

Title
Governing at Machine Speed: An Adaptive Intelligence Architecture for Real-Time AI Policy Enforcement
Author
Sandeep Bokkasam, B. Durgalakshmi
Publication
arXiv
Date
Tuesday, September 15, 2026