Federal Government Struggles Financially and Intellectually in Face of AI Advancements
AI is galore in workplaces worldwide, but not so in the U.S. federal government. A fresh report from Fedscoop ponders the federal government's inability to keep pace with the rapidly advancing AI landscape. A comprehensive analysis of various federal agencies reveals that D.C. struggles with insufficient funds and inadequate talent for AI training.
In a spine-tingling moment last year, President Joe Biden was petrified watching Mission: Impossible-Dead Reckoning Part One, leading him to issue an executive order urging tech companies to manage AI responsibly. The White House also requested federal agencies to submit reports detailing their AI plans, risk mitigation strategies, and barriers to widespread adoption of AI by September.
Most federal agencies responded, and Fedscoop has compiled all the reports in one convenient location. Remarkably, among the twenty-nine submitted reports, over a dozen mentioned data challenges, six pointed to a shortage of AI-trained employees, and six cited inadequate funding as obstacles to their AI initiatives.
The Department of Energy, responsible for the nation's nuclear arsenal, lamented about security concerns surrounding cloud services hindering its AI adoption. Additionally, it faces a scarcity of graphics cards.
The DOE's report on AI adoption elaborated on this predicament, stating, "The IT infrastructure barrier extends beyond the serverless CSP services to the availability and timeliness of securing virtual machines with the requisite Graphics Processing Unit (GPU) hardware to develop, train, manage, and deploy advanced AI models. This challenge is industry-wide; however, it will impact the rollout and adoption of more advanced customized use cases that require dedicated GPU hardware."
The widespread data problem is a glaring issue. Many federal agencies have endured for decades, with employees rotating in and out. Old technological systems persist due to their functional lifespan and not proactive upgrades.
This was painfully clear with America's nuclear weapons systems, which relied on enormous eight-inch floppy discs to run software governing nuclear command and control until 2019. When Colin Powell became Secretary of State under George W. Bush in 2001, he found his office filled with outdated computers from a defunct company in the pre-internet era.
The mass adoption of AI is forcing D.C. to face similar technological challenges all at once across all its departments. The data problem, echoed throughout the reports, is a manifestation of this haphazard system establishment. Centralizing and securing data for internal AI training is a pressing concern for many agencies, as off-the-shelf solutions are often not secure enough for government work.
Another recurring theme was a lack of understanding and apprehension towards AI among the workforce. The Nuclear Regulatory Commission acknowledged its employees' interest in AI but expressed reservations as well as a general lack of AI knowledge. To address this, the agency aims to bolster effective change management to empower its workforce to fully leverage AI as it is gradually introduced.
Funding constraints also posed problems for many agencies. The NRC highlighted its resource limitations, stating, "AI use cases compete for funding and staffing with other important priorities at the Bank including non-IT investments in core EXIM capabilities, cyber security, and other use cases in our modernization agenda."
These federal agencies face an uphill battle to keep up with the rest of the nation in terms of AI adoption. Reminiscent of Powell's dilemma in 2001 when he had to purchase computers for his entire office, he estimated the cost at 44,000 machines. "Don't ask me how I got the money, because I won't tell you," he said at an IT symposium in 2019.
- The White House's request for federal agencies to submit AI plans by September signifies a growing need for AI adoption within the U.S. federal government.
- A common obstacle mentioned in the submitted reports is the data challenges that federal agencies face, which are a manifestation of their haphazard system establishment.
- Centralizing and securing data for internal AI training is a pressing concern for many agencies, as off-the-shelf solutions are often not secure enough for government work.
- Funding constraints pose problems for many agencies, with the NRC highlighting that AI use cases compete for funding with other important priorities, such as cybersecurity and core EXIM capabilities.