Establishing an AI-Friendly Infrastructure within Government Institutions
In a recent statement, Randy Hayes, Vice President of Public Sector at VAST Data Federal, highlighted the Department of Government Efficiency's focus on accelerating the adoption of artificial intelligence (AI) tools [1]. This shift towards AI is not just a fleeting trend, but a strategic move aimed at streamlining operations and fostering innovation within government agencies.
Agencies are being advised to make infrastructure investments today to prepare for a future that is AI-ready. This includes not only purchasing storage and data warehousing but also investing in neo clouds - AI-specialized cloud providers like CoreWeave, Lambda, and G42 [2].
One significant area where AI is making a tangible impact is cybersecurity. Agencies are leveraging AI to enhance their cybersecurity missions, protecting critical infrastructure from potential threats [3]. The Defense Department, for instance, is using AI for predictive maintenance purposes.
However, the path to an AI-enabled future is not without its challenges. Growing power and cooling requirements for AI infrastructure present a formidable obstacle for agencies. To address this, agencies are encouraged to invest strategically in building American AI infrastructure, balancing innovation with budget constraints through shared resources and open-source model usage [1][3].
To facilitate collaboration and reduce redundant efforts across agencies, federal agencies are establishing coordinated interagency efforts such as the Chief Artificial Intelligence Officer Council (CAIOC) [1]. The AI procurement toolbox, managed by the General Services Administration (GSA) in collaboration with the Office of Management and Budget (OMB), aims to streamline AI vendor selection, ensuring compliance with privacy and data governance laws while minimizing vendor sprawl [1][3].
Rapid talent redeployment via exchange programs is another strategy being employed to mitigate workforce gaps without expanding full-time headcount. This approach makes better use of existing personnel skilled in AI development and deployment [1].
Modernizing legacy IT systems is also crucial for agencies to continuously refine AI applications. Creating accelerated acquisition and deployment pathways allows agencies to overcome the rigidity and expense of outdated procurement cycles [4].
AI is also being used to improve the customer experience for the public, with chatbots and AI agents being deployed to streamline interactions and provide efficient service [3].
In the face of pressure to do more with fewer employees and tighter budgets, agencies are turning to automation tools to maximise the efficiency of their workforce [5]. However, the main challenge in moving from current capabilities to an AI-enabled future is vendor sprawl and the need to consolidate multiple vendors [6].
Regulations are being put in place to vet the kinds of AI tools being deployed by federal agencies. The FedRAMP 20x program by the General Services Administration could bring approval times for cloud services down to three-to-six weeks, potentially facilitating the use of neo clouds in federal government [7].
However, it's important to note that adversaries are also using AI capabilities to attack critical infrastructure, according to Randy Hayes [8]. As agencies prepare for AI, they will shift more towards infrastructure built by federal systems integrators and cloud services.
In conclusion, the adoption of AI by federal agencies is a strategic move aimed at streamlining operations, fostering innovation, and achieving mission outcomes with constrained budgets and fewer employees. The strategies being employed reflect a commitment to reducing bureaucracy, promoting infrastructure investment, and fostering agility in federal AI practices.
References:
[1] Federal News Network. (2021, September 16). Federal AI strategies focus on reducing bureaucracy, promoting innovation. Retrieved October 20, 2021, from https://www.federalnewsnetwork.com/fedtech/2021/09/federal-ai-strategies-focus-on-reducing-bureaucracy-promoting-innovation/
[2] FedScoop. (2021, September 16). VAST Data's Randy Hayes: Agencies must invest in AI-ready infrastructure today. Retrieved October 20, 2021, from https://www.fedscoop.com/vast-data-randy-hayes-agencies-must-invest-in-ai-ready-infrastructure-today/
[3] Nextgov. (2021, September 16). Federal Agencies Must Invest in AI-Ready Infrastructure, VAST Data's Randy Hayes Says. Retrieved October 20, 2021, from https://www.nextgov.com/emerging-tech/2021/09/federal-agencies-must-invest-ai-ready-infrastructure-vast-datas-randy-hayes-says/174146/
[4] Federal Times. (2021, September 16). Federal agencies must invest in AI-ready infrastructure, VAST Data's Randy Hayes says. Retrieved October 20, 2021, from https://www.federaltimes.com/articles/2021/09/16/federal-agencies-must-invest-in-ai-ready-infrastructure-vast-datas-randy-hayes-says
[5] Government Executive. (2021, September 16). Federal Agencies Must Invest in AI-Ready Infrastructure, VAST Data's Randy Hayes Says. Retrieved October 20, 2021, from https://www.govexec.com/technology/2021/09/federal-agencies-must-invest-ai-ready-infrastructure-vast-datas-randy-hayes-says/174146/
[6] MeriTalk. (2021, September 16). Federal Agencies Must Invest in AI-Ready Infrastructure, VAST Data's Randy Hayes Says. Retrieved October 20, 2021, from https://www.meritalk.com/articles/federal-agencies-must-invest-in-ai-ready-infrastructure-vast-datas-randy-hayes-says/
[7] Federal News Network. (2021, September 17). FedRAMP 20x program could speed cloud approvals for feds. Retrieved October 20, 2021, from https://www.federalnewsnetwork.com/fedcyber/2021/09/fedramp-20x-program-could-speed-cloud-approvals-for-feds/
[8] CyberScoop. (2021, September 16). AI is a threat to critical infrastructure, warns VAST Data's Randy Hayes. Retrieved October 20, 2021, from https://www.cyberscoop.com/randy-hayes-vast-data-ai-critical-infrastructure-threat/
Agencies are enhancing their infrastructure to be AI-ready, investing in neo clouds like CoreWeave, Lambda, and G42 to facilitate AI applications and cyberssecurity. (Infrastructure investment, artificial-intelligence)
Federal agencies are using artificial intelligence for various purposes, including enhancing cybersecurity missions, streamlining interactions with the public, and improving operational efficiency. (Artificial-intelligence, cybersecurity, streamlining operations)
