U.S. Presses Meta to Agree to A.I. Reviews
The U.S. government is pressuring Meta to agree to independent reviews of its artificial intelligence systems, citing national security concerns. Officials are pushing for third-party assessments of Meta's open-source AI models before they are publicly released, as part of broader efforts to regulate the rapidly advancing technology.
Background
- Meta (Facebook's parent company) is developing powerful AI models (e.g., Llama) that it releases as open-source — meaning anyone can download, modify, and reuse them freely. This approach has made Meta a central player in the AI race alongside OpenAI, Google, and Microsoft.
- The U.S. government, through agencies like the National Institute of Standards and Technology (NIST) and the AI Safety Institute, is increasingly pushing for mandatory safety testing of advanced AI models before they are deployed — especially when they are open-source and therefore harder to control after release.
- The core tension: Meta argues that open-source AI fosters innovation and transparency; critics (and now the U.S. government) argue that powerful models with no safety guardrails could be used for cyberattacks, bioweapons design, or disinformation at scale. The article reports that the Biden administration is pressing Meta to submit its latest models to formal government review processes, marking a significant escalation in AI regulation efforts.