The OMB’s Policy to Manage Federal Use of AI: What You Need to Know

On 28 March 2024, the White House Office of Management and Budget (OMB) issued its policy on Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence. The policy directs federal agencies to further AI innovation and governance while establishing and managing the risks arising from the use of AI, especially those that affect the safety and rights of the American public.

How does the OMB policy on AI align with other US initiatives?

The OMB’s policy is aligned with Biden’s Executive Order 14110 on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, and marks 180 days since the executive order was issued, with all 150-day actions required by federal agencies having been completed.

Specifically, the Governance, Innovation, and Risk Management Policy administers a key component of the Executive Order to strengthen AI safety and develop AI responsibly. Section 10 of the Executive Order deals with ‘Advancing Federal Government Use of AI’, under which the OMB is mandated to convene and chair an interagency council to coordinate the use of AI in agencies, and issue guidance to agencies on the responsible use of AI, innovation, and risk management. This new policy by the OMB falls under their scope to issue guidance.

The policy represents a push to strike the correct balance between responsibility and innovation, and entrusts the federal government to securely develop, test, procure, and integrate AI administration-wide. It also builds on the National Institute of Standards and Technology’s (NIST) AI Risk Management Framework and the White House’s Blueprint for an AI Bill of Rights, and will make federal agencies accountable for both the development and risk management of AI.

What actions does the OMB policy on AI require from federal agencies?

In line with executive order 14110, the OMB policy outlines several AI governance, innovation, and risk management actions for federal agencies, including specific minimum risk management practices.

Address Risks from the Use of AI

Agencies must implement safeguards when using AI in a way that could impact the rights and safety of Americans. These safeguards must be implemented by 1 December 2024 and include compulsory actions to transparently inform the public on how the government is using AI and monitoring and mitigating any discriminatory impacts. Agencies must stop using any AI system if they cannot implement these safeguards, unless they can prove that using such a system is necessary for agency operation or a cease would increase overall risks to rights and safety.

Agencies must also manage risks specifically pertaining to the procurement of AI, and the OMB will take specific action later this year to ensure that procurement contracts align with the new policy.

Expand Transparency of AI Use

Agencies must follow detailed instructions to publicly report:

• Annual inventories of their AI use cases
• Metrics about AI use cases that are not publicly declared due to their sensitivity
• Any AI system exempted from complying with OMB policy, along with the associated reasons
• Government-owned AI code, models, and data, where releasing such information does not pose any additional risks

Advance Responsible AI Innovation

AI technology is already being used by various agencies to address societal challenges – the Center for Disease Control has been using AI to predict the spread of disease, while the National Oceanic and Atmospheric Administration is using AI to forecast extreme weather more accurately. The use of generative AI, which is being currently being piloted by several agencies using AI chatbots to improve customer experience for instance, will further this innovation and create more opportunities for innovation.

The new OMB policy will remove barriers to federal AI innovation by encouraging agencies to continue to experiment with generative AI with appropriate safeguards in place.

Grow the AI Workforce

Agencies are strengthening their AI talent by increasing personnel and upskilling workers. Additional budget allocations have been made for government-wide AI training, and the Administration has committed to hiring 100 AI professionals by summer 2024 and will be expanding AI roles across government.

Strengthen AI Governance

Federal agencies are required to designate Chief AI officers to coordinate the use of AI across agencies, and establish AI Governance Boards to govern the use of AI within agencies. These governance boards will be chaired by the Deputy Secretary or equivalent, and must be established by 27 May 2024.

What other measures are being undertaken by the Administration?

Alongside the new OMB policy, the government has also introduced the following actions to promote responsible AI use:

• The OMB will issue a Request for Information (RFI) on the responsible procurement of AI in government
• Agencies will increase reporting on the 2024 Federal AI Use Case Inventory

Prioritize AI Compliance with Holistic AI

The OMB’s new policy to drive federal accountability of AI represents a growing landscape of greater vigilance and compliance.

Schedule a demo with our experts to find out how our AI Governance Platform can help you stay on top of AI legislation, regulation, guidance, and more around the world.