Innovation: Artificial Intelligence Talent (Public Sector)

Sector: 23.7% of GDP (Federal Gov’t) 
Estimated Cost: $1B / ~0% of federal budget
Political Challenge: Low  

I. Overview

Artificial intelligence technology promises to fortify our national security and put unprecedented capabilities at the fingertips of public agencies. But the U.S. public sector has a technical talent problem, putting the technology partially out of reach and our security, society, and growth at risk. Outdated federal hiring practices and policies are to blame, and updating them is essential for bringing government into the modern age.  

II. Problem 

More than half of agency staff feel their agencies are underequipped to “build, manage, or procure” artificial intelligence.¹ 

With a strong and permanent base of artificial intelligence talent, the federal government could:  

• plan and pursue preliminary use cases for AI, which may involve managing highly technical contracts 

• advance smart rules for the technology’s use in various regulatory domains, especially to head off existential risk 

• equip itself for a new world of cyberdefense 

• make better use of programs that bring AI experts into government temporarily 

Agencies are aware of these needs, planning hundreds of AI roles or “AI-enabling” roles (the more common). The Department of Defense is the largest potential employer, seeking to hire 9,000 people in AI-relevant roles by FY 2026.²  

Outdated hiring policies undermine this talent base. Attracting and retaining any technical talent is difficult due to pay gaps with the private sector. Many top professionals participate in programs that offer time-limited roles in government instead. But hiring these experts is a challenge even when they do want to work in the public sector, because the conventional, “competitive” hiring process puts low weight on demonstrated expertise.³  

The Office of Personnel Management (OPM) extended wide-ranging excepted service and direct hire authorities for AI roles in 2023 and 2024. It has also encouraged cross-agency pooled hiring, which promises to reduce costly redundancies, and offered guidance to help agencies do skills-based hiring.⁷ This showed some success in the AI Talent Surge, which brought 250 AI professionals into government.⁸  

Agencies are emerging as a bottleneck. The full OPM hiring toolkit has gone underutilized due to risk-aversion and compliance fears from agency HR offices.⁹ This is itself partly a result of OPM’s and agency watchdogs’ typically conservative approach to hiring¹⁰ — a culture of slow approvals and frequent rejections when agencies step outside the lines — and some real and imagined compliance risks that have become agency conventional wisdom.¹¹  

III. Recommendations 

Federal talent is a practice problem, not (just) a policy one. To build a stronger technical workforce, OPM must (counterintuitively) create a new culture of greater agency autonomy. This means educating agencies about the tools at their disposal and exercising a lighter touch when reviewing agency hiring actions. While this expanded autonomy itself should cause shifts in their hiring strategies, agencies must be offered additional incentives by OPM and political appointees to use it to its fullest.

To complement these efforts, Congress can expand rotational programs, like Presidential Innovation Fellows, so that top talent can move back and forth between the public and private sectors with more ease. At a higher level of ambition, Congress could revamp the conventional hiring process to be more skills-based.  

IV. Risks and Politics 

The bulk of recommended changes can be conducted without much controversy within the executive branch. Legal and political opposition from groups representing veterans (who are privileged by the conventional process) may present a challenge.