This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| 1 minute read

New OMB Policy Sets Guidelines for Managing Risks Associated with Federal Procurement of AI

On Thursday, March 28, 2024, The Office of Management and Budget (OMB) published a new policy—Advancing Governance, Innovation, and Risk Management for Agency Use of Artificial Intelligence (AI)—directing agencies to seize the opportunities AI presents while also managing its risks. Unsurprisingly, OMB’s policy extends to the risks associated with procuring what it dubs as rights-impacting AI and safety-impacting AI. For those of us in the Government Procurement arena, this policy signals that significant change is afoot for AI contractors.

Based on the OMB’s recommendations for agency AI procurement, at a minimum, federal contractors selling AI services or systems to the government should expect to see contractual provisions requiring:

  • Compliance with privacy, confidentiality, IP, cybersecurity, human rights, civil rights, and civil liberties laws, regulations, and policies;
  • Transparency;
  • Adequate documentation assessing your AI’s capabilities, limitations, and intended uses; 
  • Continuous improvement of procured AI; 
  • Post-award monitoring of the AI; 
  • The Government to maintain sufficient rights to the data and any improvements to the data; 
  • Interoperability;
  • The protection of Federal information used in the development and operation of AI products and services;
  • Restrictions on any contractor use of the data to train or improve your commercial AI offerings without the express permission of the agency; 
  • Restrictions on the use of agency data to train your AI systems; 
  • Appropriate consent and lawful collection of data that is free from unwanted bias;
  • Documentation or test results validating the accuracy, reliability, and validity of the AI’s abilities; 
  • Adequate testing and safeguards; 
  • Test results;
  • Reliable labeling of AI generated or modified content;
  • Incorporation of relevant NIST standards; and 
  • Documentation showing your methods for improving environmental efficiency and sustainability of the AI.  

Think this is a long list? Unfortunately, these requirements are just the beginning. There’s no question that AI is going to revolutionize the landscape of government contracting, but that revolution will include boundaries. OMB’s policy is just the beginning. 

OMB’s guidance sets up several “binding requirements to promote the safe, secure and responsible use of AI by our federal government.” - Vice President Kamala Harris

Tags

ai, defense and national security, government relations, ip, privacy cybersecurity, regulatory