Federal Agencies to Leverage Cloud for Artificial Intelligence R&D
Published: December 21, 2020
Using cloud for federal R&D related to artificial intelligence could boost market size.
- The National Science and Technology Council is advising that federal artificial intelligence R&D do more to leverage commercial cloud computing.
- Recommendations made by the council are likely to be supported by the incoming administration.
- Increased federal investment in artificial intelligence R&D could mean a boost in federal cloud spending.
On November 17, 2020, the White House Select Committee on Artificial Intelligence, a part of the National Science and Technology Council, published an intriguing paper titled “Recommendations for Leveraging Cloud Computing Resources for Federally Funded AI Research and Development.” The paper offered a series of ideas concerning a subject that we here at Federal Market Analysis discuss relatively often, this being the fact that cloud computing will eventually become the primary platform for delivering artificial intelligence capabilities.
The committee starts from the premise that artificial intelligence will be a major economic driver for the United States. Researchers working on nascent AI projects can in the meantime leverage the massive computing power offered by commercial cloud companies to help develop ever more sophisticated AI capabilities. To that end, the committee offered four recommendations for how federal agencies “can better enable the use of cloud computing resources for federally funded AI R&D.
The incoming presidential administration has already expressed support for increasing AI R&D funding, a fact that suggests it will sympathize with the objectives outlined in the committee’s paper and support heavier investment in the years to come. Should it do so, the funds flowing into AI R&D should also benefit some commercial cloud infrastructure providers.
Here are the committee’s recommendations:
- Launch and support pilot projects that help identify and explore the advantages and challenges associated with the use of commercial clouds in conducting federally funded AI research.
- Improve education and training opportunities to help researchers better leverage cloud resources for AI R&D.
- Catalog best practices in identity management and single sign-on strategies to enable more effective use of the variety of commercial cloud resources for AI R&D.
- Establish and publish best practices for the seamless use of different cloud platforms for AI R&D.
Pilot Projects – These will not appear everywhere. Look for potential work associated with these initiatives at agencies such as the National Science Foundation, Department of Energy, National Aeronautics and Space Administration, and Department of Defense. Keep an eye on the NSF in particular as it is being positioned as a leader in AI R&D. The DOD will be interested simply by virtue of its vast R&D enterprise and growing comfort with commercial cloud computing.
Security Best Practices – Implementing identity management and single sign-on capability is an ongoing process for most federal agencies. It is also based on commercial best practices, suggesting the possibility of related consulting work. At its heart, the best security “best practice” today is zero trust so keep tabs on particularly the DOD’s efforts to implement such an architecture as it could provide a roadmap for other agencies to follow.
Cloud Platform Best Practices – Cloud lock-in is already a concern for federal agencies, presenting a Hotel California-like situation where agencies can check-in, but never leave. Advancing AI R&D priorities could end up forcing a change to this situation if interoperability becomes a serious need for high-profile projects.
Summing up, here is Deltek’s latest forecast projects the addressable federal cloud market will grow by FY 2022 to just shy of $8B. Federal Market Analysis arrived at this number prior to the publication of the NSTC committee’s paper on AI R&D, however, given the interest in using commercial cloud for AI R&D, our projection might be conservative as federal R&D attracts significant funding in the years to come.