Robyn N. Burrows and Sara N. Gerber ●
Last month, the Office of Management and Budget (“OMB”) issued a memorandum directing federal agencies to adopt artificial intelligence (“AI”) and advance its use to inform and carry out agency actions. OMB’s new policy addresses three main areas it views as necessary for responsibly deploying AI in agency decision-making: (1) strengthening AI governance; (2) advancing AI innovation; and (3) managing risks from the use of AI. With OMB encouraging the use of AI to streamline agency actions wherever possible, government contractors can also expect to see AI increasingly used in the procurement process.
AI Governance
OMB directed agencies to designate a Chief AI Officer whose responsibilities will include coordinating agency use of AI, developing a workforce with the skillsets necessary for implementing AI, and “identifying and prioritizing appropriate uses of AI that will advance both their agency’s mission and equitable outcomes.”
The Chief AI Officer is also tasked with ensuring that AI code and the data used to develop and test AI are inventoried and shared in data repositories. That individual must also prepare and submit annually to OMB an “AI use case inventory” documenting instances in which AI is used to address a particular need. For example, the Department of State’s (“DOS”) AI Inventory includes a bot that it developed “to automate the data entry in the Federal Procurement Data System” which the State Department reports has reduced the burden on the agency’s procurement staff and improved compliance on DATA Act reporting.
AI Innovation
OMB views AI as having the potential to improve efficiency across the federal government and envisions its use to increase access to government services, address the climate crisis, protect democracy, improve public health, and grow economic competitiveness.
To this end, OMB directed agencies to remove the barriers to the responsible use of AI with the objective of achieving “AI maturity.” Each agency must propose a plan for developing the IT infrastructure necessary to build, test, and maintain AI applications. Agencies must also develop the infrastructure necessary to share and curate internal and external data for use in training and operating AI. OMB further directed agencies to assess potential beneficial uses of generative AI to their missions.
OMB views AI as a tool that must be open, shared, and reused. Agencies “must proactively share custom-developed” AI code for “AI applications in active use” and must “maintain that code as open source software on a public repository.” Agencies are also required to publicly share the data used to develop and test AI.
Managing Risks
OMB acknowledged risks associated with AI, which it categorized as either “safety-impacting” (e.g., impacting human life, climate, critical infrastructure, etc.) or “rights-impacting” (e.g., impacting civil rights, equal opportunities, access to critical government resources/services, etc.). OMB prescribed minimum risk management practices that agencies must follow before using AI, which include completing an AI impact assessment documenting the intended purpose for the AI; its expected benefit (e.g., cost reduction); potential risks of using AI, including its impact on equity and fairness; and an assessment of the quality of the data used in the AI’s development and testing. Agencies must also test the AI to ensure it will work in its “intended real-world context” and must obtain an independent evaluation of the AI (through an oversight board or other agency office) to ensure the benefits outweigh the risks.
For federal procurements of AI, OMB directed agencies to promote competition and interoperability to avoid improperly entrenching incumbents. Similarly, OBM directed agencies to ensure sufficient government data rights to avoid vendor lock-in and to allow the government’s continued design, development, testing, and operation of AI. OMB also issued a request for information seeking input on the federal government’s procurement of AI, including proposed contract terms to address data rights issues.
Use of AI in Government Contracting
With OMB’s direction to agencies to remove barriers to the use of AI to improve the efficiency of services and operations, government contractors can expect agencies to increasingly implement AI in acquisitions and contract management. Some agencies are already beginning to do so. For example:
- The General Services Administration (“GSA”) developed CALI, an automated machine learning tool that evaluates vendor proposals against solicitation requirements to support the source selection process. GSA’s AI Inventory explains that CALI analyzes proposals in four “key” areas: “format compliance, forms validation, reps & certs compliance, and requirements compliance.”
- DOS has been using a bot developed to automate closeout reminders for federal assistance grants nearing the end of the period of performance.
- In FY 2023, the Department of Homeland Security (“DHS”) conducted a pilot program to evaluate an AI market research tool to identify prospective contractors capable of performing a requirement. DHS reported that without the AI tool, DHS acquisition employees spent more than 10 hours of work identifying vendors capable of performing a requirement. The use of the AI tool reduced that time to 2.5 hours.
- Several other AI tools are intended to detect fraud, including fraud prevention system models that capture Medicare administrative and claims data to identify potential cases for future investigation by the Centers for Medicare & Medicaid Services. DOS is also developing “Supply Chain Fraud and Risk Models” to identify anomalous activity within the Integrated Logistics Management System that could represent potential fraud or malfeasance.
Contractors may also be able to use AI to their benefit, for example, to help validate compliance with regulatory requirements, determine contracts they are eligible to bid on, or to assist in preparing proposals. Simplifying the contracting process and reducing associated costs could, in turn, help encourage more companies, especially small businesses, to become more willing to bid on contracts, increasing competition and innovation.