White House AI Framework: Security Risks & Closed Models

The White House is preparing an AI framework to review security risks. Officials met with top artificial intelligence companies to discuss this new approach. Here's the thing: The government plans to review only "closed" AI models, which do not publish their underlying code. These are made by companies like Anthropic and OpenAI. What's interesting is the administration does not plan to review "open source" AI models. These models have code available to the public. However, this could change as the technology develops. This framework is an effort to formalize oversight of leading AI labs. It follows months of private discussions with major tech companies, including representatives from Anthropic, OpenAI, Microsoft, Meta, Google, and Nvidia. The decision to exclude open-source models could benefit companies trying to compete with current leaders. However, it also delays addressing potential threats from powerful open-source models, especially those from Chinese firms. The bottom line: This framework will likely influence the global AI race and the rollout of new AI technologies.