The Federal Government Is Falling Behind on AI Skills, But Here Is How It Can Catch Up

AI offers opportunities to improve federal service delivery. Applications for AI in government include automating tasks and data entry, enhancing virtual assistants and chatbots, optimizing resource allocation and budgeting, and even assisting in coding and website development. A Deloitte study nearly five years ago found that AI could save around 30 percent of government employees’ time in a matter of years. Unfortunately, the federal government does not have the skilled workforce to take advantage of AI’s potential and is acting too slowly to remedy this situation. As the U.S. government seeks to accelerate its adoption of AI, federal agencies should fully evaluate AI literacy and skills in their respective workforces so that they can address critical skills gaps before falling even further behind in AI adoption.

Many federal agencies are lagging in both planning for and adoption of AI, despite good intentions a few years ago from both Congress and the Trump administration. For example, many agencies have struggled to maintain accurate and up-to-date public records of how they are using AI, despite a requirement to do so in a December 2020 executive order (EO) from the Trump Administration. The Office of Personnel Management (OPM) has also failed to fulfill three of the four tasks required of it in the AI in Government Act of 2020, legislation intended to accelerate the federal government’s adoption of AI. Likewise, the legislation directed the Office of Management and Budget (OMB) to issue a memo to federal agencies regarding the policies for acquiring AI solutions and products, but that memo is still forthcoming.

There are two steps the federal government should take to address AI skills in its workforce. First, OPM should complete the rest of the tasks Congress assigned to it in the AI in Government Act. Specifically, OPM should establish or revise job series where the primary duties are related to AI, determine an estimate of existing federal positions related to AI by agency, and prepare a two- and five-year forecast on the number of AI-related jobs federal agencies will need. So far, OPM has only shared a memo listing both general and technical competencies for AI-related work. This memo is a helpful reference for federal agencies looking to include such competencies in a job posting, but agencies need a thorough understanding of AI literacy and skills in their workforces so they can focus on creating new AI-related positions as needed, recruiting new talent with AI skills, and developing AI upskilling programs for existing employees.

Second, federal agencies should cultivate continuous learning environments where new and current employees can quickly acquire technical skills related to digital technologies. Internal classes, certification programs, or even a fully-fledged digital skills academy would not only support necessary AI skills development but also the federal government’s broader digital transformation efforts. After all, digital isn’t going anywhere, and new technologies will continue to emerge. The federal government needs a workforce that can acquire, implement, and use digital technologies effectively for any federal digital adoption initiatives to be successful.

The federal government spends billions of dollars on AI technologies each year. Such spending will achieve limited impact if there aren’t enough federal employees who can build and operationalize these tools. OPM should not only comply with existing laws and mandates to get AI skills into the federal government as fast as possible, but the federal government should also consider how it can create an environment that supports continuous learning of digital skills broadly among its workers. Only then will the federal government be able to take advantage of AI’s unique potential—or the potential of any digital technology—to improve operations and service delivery.