The US government on Thursday issued new rules that require more caution and transparency from federal agencies using artificial intelligence. They say these are needed to protect the public as AI rapidly develops. But the new policy also includes provisions to encourage AI innovation in government agencies when the technology can be used for the public good.
The US hopes to emerge as an international leader with its new government AI regime. Vice President Kamala Harris said at a news briefing ahead of the announcement that the administration plans for the policy “to serve as a model for global action.” She said the US “will continue to call on all countries to follow our lead and put the public interest first when it comes to government use of AI.”
The new policy from the White House Office of Management and Budget will guide the use of AI within the federal government. It requires more transparency about how the government uses AI and also calls for more development of the technology within federal agencies. Under the policy, the government is seeking to strike a balance between mitigating the risks of deeper use of AI – the extent of which is unknown – and using AI tools to solve existential threats such as climate change and disease.
The announcement adds to a series of steps by the Biden administration to embrace and limit AI. In October, President Biden signed a sweeping executive order on AI that would promote the government’s expansion of AI technology, but also require those who create large AI models to provide the government with information about their activities, in the interest of national security .
In November, the US, along with Britain, China and members of the EU, signed a declaration that recognized the dangers of rapid progress in AI but also called for international cooperation. Harris unveiled a non-binding statement on military use of AI the same week, signed by 31 countries. It sets up rudimentary guardrails and calls for the deactivation of systems that engage in “unintended behavior.”
The new policy for the use of AI by the US government, announced Thursday, calls for agencies to take several steps to prevent unintended consequences of AI implementations. To start, agencies must verify that the AI tools they use do not endanger Americans. For example, if the Department of Veterans Affairs wants to use AI in its hospitals, it must verify that the technology does not produce racially biased diagnoses. Research has found that AI systems and other algorithms used to diagnose or decide which patients receive care can reinforce historical patterns of discrimination.
If an agency cannot ensure such safeguards, it must stop using the AI system or justify its continued use. U.S. agencies face a December 1 deadline to comply with these new requirements.
The policy also calls for greater transparency about government AI systems, requiring agencies to release government AI models, data, and codes, as long as releasing such information does not pose a threat to the public or government. Agencies must publicly report every year how they use AI, what potential risks the systems pose and how these risks are mitigated.
And the new rules also require federal agencies to increase their AI expertise, with each appointing a Chief AI Officer to oversee all AI used within that agency. It is a role that focuses on promoting AI innovation and guarding against its dangers.
Officials say the changes will also remove some barriers to AI use across federal agencies, a move that could facilitate more responsible experimentation with AI. The technology has the potential to help agencies assess damage after natural disasters, predict extreme weather, map the spread of disease and monitor air traffic.
Countries around the world are in the process of regulating AI. The EU voted in December to adopt the AI Law, a measure regulating the creation and use of AI technologies formally adopted it earlier this month. China is also working on it extensive AI regulations.