Biden Admin Mandates AI Safety in Federal Agencies

The Biden administration on Thursday released a new memo from the Office of Management and Budget providing guidance to federal agencies on artificial intelligence use, in a significant step aimed at ensuring safe and responsible adoption of the rapidly advancing technology.

The policies, which build on an executive order signed by President Biden in October, require agencies to assess and mitigate potential risks posed by their AI systems, such as algorithmic bias or threats to privacy and safety. By December 1st, agencies must implement safeguards to “reliably assess, test and monitor AI’s impacts on the public,” or else cease use of the technology unless it is deemed mission-critical.

The guidance also calls for “more transparency about government AI systems, requiring agencies to release government-owned AI models, data, and code,” as long as doing so does not jeopardize security. Agencies will need to publicly report each year on their AI use and the steps taken to address potential risks.

In addition, all federal agencies are now required to appoint a chief AI officer to oversee implementation and innovation of artificial intelligence. The administration aims to “grow the workforce focused on AI, including by hiring at least 100 professionals in the field by this summer.”

Vice President Kamala Harris emphasized that the policies were developed with input from a range of stakeholders and are intended to “serve as a model for global action.” “President Biden and I intend that these domestic policies will serve as a model for global action,” Harris said. “We will continue to call on all nations to follow our lead and put the public interest first when it comes to government’s use of AI.”

The guidance aims to balance risk management with encouraging responsible innovation, as AI holds potential to help agencies respond to challenges like natural disasters, disease outbreaks, and transportation issues. But concerns remain about the extent of unknown risks as the technology rapidly evolves. The policies are part of a series of actions the administration has taken to embrace AI’s benefits while restraining its dangers.