New US Federal AI Rules: A Guide for Government Suppliers
![Imak___tech_lab_with_a_green_plant_254ed861-2e97-4c02-a64f-c5f71bda2b24 - Open Opportunities](https://www.openopps.com/wp-content/uploads/2024/04/imak___tech_lab_with_a_green_plant_254ed861-2e97-4c02-a64f-c5f71bda2b24.png)
The new Office of Management and Budget (OMB) memorandum introduces two key classifications for AI systems: “rights-impacting AI” and “safety-impacting AI.” These classifications determine the minimum practices that agencies must follow when deploying AI solutions. If you’ll supply to the US government, it is crucial to understand how your AI offering aligns with these definitions and be prepared to make your case during agency assessments.
Preparing for Compliance and Reporting
Companies delivering AI solutions to federal agencies must comply with the minimum practices outlined in the OMB memorandum. These practices include conducting AI impact assessments, independent evaluations, ongoing monitoring, risk mitigation, and providing public notice and documentation. Suppliers should anticipate and prepare for these requirements, ensuring adequate documentation, guidelines, and risk management measures are in place for their AI offerings.
Interoperability, Data Rights, and Code Sharing
The OMB memorandum emphasises the importance of interoperability, data rights, and code sharing. Government suppliers should expect agencies to require AI solutions that work across multiple cloud environments and grant the government sufficient rights to data, improvements, and custom code. Additionally, agencies are encouraged to publicly share and release AI code, models, and data, potentially requiring suppliers to grant rights for public release.
Furthermore, the memorandum emphasises the importance of interoperability, data rights, and code sharing. Government suppliers should expect agencies to require AI solutions that work across multiple cloud environments and grant the government sufficient rights to data, improvements, and custom code. Additionally, agencies are encouraged to publicly share and release AI code, models, and data, potentially requiring suppliers to grant rights for public release.
As companies navigate the new federal AI rules, staying informed and monitoring agency-specific requirements is essential, especially for those with defence and national security missions. Suppliers should also be mindful of existing legal requirements related to privacy, confidentiality, intellectual property, cybersecurity, human and civil rights, and civil liberties, as well as any additional requirements for generative AI and dual-use foundation models.
Are you looking for B2B opportunities?
Want a preview of our new Open Opportunities platform tracking global government spending? Spend Network has developed a new tool, to allow easy access to global opportunities to sell to governments. Open Opportunities is the most comprehensive tool on the market for government spending worldwide.Sign up now!
Open new markets, use our in-depth data to identify expansion opportunities, partners, and excite customers.