Black-Box Access is Insufficient for Rigorous AI Audits. (arXiv:2401.14446v1 [cs.CY])

External audits of AI systems are increasingly recognized as a key mechanism
for AI governance. The effectiveness of an audit, however, depends on the
degree of system access granted to auditors. Recent audits of state-of-the-art
AI systems have primarily relied on black-box access, in which auditors can
only query the system and observe its outputs. However, white-box access to the
system's inner workings (e.g., weights, activations, gradients) allows an
auditor to perform stronger attacks, more thoroughly interpret models, and
conduct …

access ai audits ai governance art arxiv audit audits box can external governance key mechanism query state system systems