Practical Principles for AI Cost and Compute Accounting

The article proposes seven principles for AI cost and compute accounting to address regulatory challenges. It argues that policymakers use cost and compute as proxies for AI capabilities and risks, and that current technical ambiguities create loopholes. The principles aim to reduce gaming, avoid discouraging risk mitigation, and enable consistent implementation across companies and jurisdictions.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.