Practical Principles for AI Cost and Compute Accounting
This addresses a policy and regulatory challenge for AI governance, but it is incremental as it builds on existing regulatory frameworks without introducing new technical methods.
The paper tackles the problem of technical ambiguities in AI cost and compute accounting used for regulation, proposing seven principles to reduce gaming, avoid disincentivizing risk mitigation, and enable consistent implementation.
Policymakers increasingly use development cost and compute as proxies for AI capabilities and risks. Recent laws have introduced regulatory requirements for models or developers that are contingent on specific thresholds. However, technical ambiguities in how to perform this accounting create loopholes that can undermine regulatory effectiveness. We propose seven principles for designing AI cost and compute accounting standards that (1) reduce opportunities for strategic gaming, (2) avoid disincentivizing responsible risk mitigation, and (3) enable consistent implementation across companies and jurisdictions.