In contrast to the EU AI Act, President Biden’s Executive Order and the earlier “Blueprint for an AI Bill of Rights” do not prohibit the use of AI and do not have enforcement mechanisms. Instead, they take a rights-based approach to regulation.
However, before releasing advanced AI systems to the public, the most recent Order “requires” that their safety test results and other information be shared with the US government. It likewise summons the Safeguard Creation Act and requires organizations creating establishment model that represents a serious gamble to public safety to inform the public authority while preparing the model and offer the consequences of security tests.
In addition, before AI systems are made available to the general public, the National Institute of Standards and Technology (NIFT) has been tasked with creating stringent standards and tools for testing, evaluating, verifying, and validating them. Notably, the Order directs agencies to combat algorithmic discrimination in the criminal justice system and urges Congress to pass a data privacy law.