tech
White House Finalizes AI Framework Behind Closed Doors
The White House said it met its deadline to establish a voluntary framework for evaluating advanced AI models — but it won't say what the framework contains.
TL;DR
- The White House met its deadline to establish a framework for evaluating advanced AI models.
- Details of the framework, including its contents and implementation timeline, are not publicly disclosed.
- The framework aims to provide AI developers with a structure for government engagement on model evaluation.
- It outlines requirements for confidentiality, cybersecurity, insider-risk, intellectual-property protection, and nondisclosure for government access to models.
- The process for assessing advanced cyber capabilities of AI models will be classified.
- The threshold for which models are covered by the order will be shared only with AI developers and researchers as appropriate.
- A staff-level meeting with companies is scheduled to review the framework.