The White House Office of Management and Budget (OMB) on March 28, 2024, published a memorandum with the subject “Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence.” This memorandum provides guidance on the usage and risk management of artificial intelligence (AI) in Government and Federal agencies.
What is Artificial Intelligence?
The publication broadly defines Artificial Intelligence (AI) as any artificial system that performs tasks under varying and unpredictable circumstances without significant human oversight, or that can learn from experience and improve performance when exposed to data sets. An artificial system developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action are also considered as AI.
Requirements for Federal Agencies in the Usage of AI
Strengthening AI Governance:
The memorandum advocates for the strengthening of AI governance within agencies and requires them to designate a Chief Artificial Intelligence Officer (CAIO) who will be responsible for coordinating the use of AI in the agency, promoting adoption, and managing risks. This section promotes a strengthening of posture in AI governance within government organizations.
- Designating Chief AI Officers (CAIO): CAIO must be qualified and have the necessary skills, training and expertise to perform the role as the lead in coordinating, innovating and risk management of the organization’s use of AI.
- Convening Agency AI Governance Bodies: The memorandum also requires each agency to convene an AI Governance Body within 60 days.
- Compliance Plans: Each agency is mandated to submit to the Office of Management and Budget (OMB) and post publicly the agency’s compliance plan in the use of AI.
- AI Use Case Inventories: Covered agencies are required to inventory each AI use case individually.
- Reporting on AI Use Cases Not Subject to Inventory: AI use cases on non-covered agencies such as the Department of Defense are also required to report or release aggregate metrics about their AI use case.
Advancing Responsible AI Innovation:
The memorandum also provides guidelines for the responsible use of AI in government agencies. Agencies are mandated to improve their ability to leverage and use AI to benefit the public while identifying the limitations and risks. Agencies are asked to use existing internal capabilities to support responsible AI innovation and explore partnership options to help scale AI innovation and usage.
- Data: Government agencies are required to develop adequate secure infrastructure to support, share, curate, and govern agency AI data.
- Cybersecurity: Agencies are mandated to have adequate cyber security measures for AI in places that are routinely updated to monitor threats from AI use.
- Generative AI: They are also mandated to assess the potential benefits of using generative AI to streamline internal processes as well as establish safeguards and oversight mechanisms to allow for generative AI to be used safely within the agency.
Managing Risks from the Use of Artificial Intelligence
The memorandum also mandates agencies to review existing policies, procedures and officials that are in place to protect the agency’s information and systems and adapt them to address the risk of the use and implementation of AI.
- Implementation or Risk Management Practices and Termination of Non-Compliant AI: Agencies must implement minimum best practices to minimize the risk and are mandated to stop using AI in their operations.
- Certification and Publication of Determinations and Waivers: Agencies must annually certify the ongoing validity of the determinations and waivers to stay compliant of the memorandum.
Conclusion
The memorandum sets the tone and provides guidelines for Government and federal agencies on the responsible use of artificial intelligence. This also paves the way for greater adaptation and accountability of AI usage in Government and Federal agencies.
