Hello, I’m Ylli Bajraktari, President of the Special Competitive Studies Project. At the AI+ Expo, the Task Force on AI and the Future of Work released its Preliminary Findings Report — the product of months of research, expert testimony, and bipartisan deliberation co-chaired by Senators Mike Rounds and Mark Warner, in partnership with NVIDIA.
The opportunity is real: AI has the potential to raise living standards, expand economic opportunity, and strengthen American competitiveness in ways we haven’t seen in generations. But getting there starts with an honest picture of where we stand. Below are our preliminary findings — the diagnostic foundation for the national action to come.
AI and the Future of Work: From Preliminary Findings to National Action
At this year’s AI+Expo, one theme surfaced repeatedly across conversations between policymakers, technologists, educators, and workforce leaders: artificial intelligence is no longer a future disruption. It is an active restructuring force across the American economy.
That reality sits at the center of the newly released AI and the Future of Work: Preliminary Findings report from the Task Force on AI and the Future of Work — a joint effort between the Special Competitive Studies Project (SCSP) and NVIDIA. The report represents the Task Force’s first assessment of how AI is beginning to reshape labor markets, skill demand, and pathways to economic mobility in the United States.
During a featured AI+Expo conversation, NVIDIA Co-Founder Chris Malachowsky and Dakota State University President Dr. José-Marie Griffiths reflected on what this transition means for America’s workforce and long-term competitiveness.
“The creativity, judgment, empathy, critical thinking, and the ability to connect dots that are not linearly associated with each other — those uniquely human characteristics are not something a mechanical agent is going to represent. AI systems represent data. Humans, together with their digital colleagues, can become far more powerful than either alone.”
— Chris Malachowsky
Dr. Griffiths emphasized another defining characteristic of the transition: the growing importance of data itself.
“One of the characteristics of this era will be data — making sure we have good data and understand what it is telling us.”
— Dr. José-Marie Griffiths
You can watch the full conversation below.
The report’s preliminary findings outline ten central dynamics shaping the AI-powered transition now unfolding across the economy:
AI is Advancing at Unprecedented Speed and More Broadly Than Prior Technologies
AI is a general-purpose technology with an unusually rapid cycle of improvement and adoption across sectors, diffusing across the United States twice as fast as the internet. Unlike most prior general-purpose technologies, it also extends into certain types of cognitive work, reshaping not only how some tasks are performed but the foundational skills with which those tasks are performed. This combination of speed and scope increases both the scale of potential gains and the uncertainty around the potential near-term disruption.
Historically, Broad-based Technological Change Has Reshaped Work, but the Exact Direction and Full Extent of AI’s Impact is Difficult to Predict
History shows that technological revolutions simultaneously create, transform, and eliminate jobs—but with outcomes that are notoriously difficult to predict in advance. While past general-purpose technology advances enhanced productivity, created new industries, and ultimately expanded work, they also involved unpredictable and uneven adjustment periods and resulted in fundamental, permanent shifts in many occupations. As with past precedent, AI has the potential to expand the frontier of human work and result in net positive job creation, but uncertainties during the transition period are likely to create disruption for at least some workers and complicate projections of the near-term impacts.
Measuring AI’s Impact on Labor Market Requires a Fundamentally Different Data Collection Strategy
Existing labor market data is not well suited to capture the speed and the breadth of changes that AI is already driving. Key indicators such as employment and productivity are generally lagging indicators that do not capture the underlying shifts in work in real time. As a result, early effects—such as changing skill demands or hiring patterns—are difficult to observe with traditional metrics and will require new data collection and analytic methodologies.
The Impact of AI is Concentrated at the Task Level First, not at the Occupation Level
Jobs are composed of bundles of tasks, and where AI is being adopted its impact appears to be concentrated in automating or augmenting the performance of specific tasks rather than eliminating entire occupations. This leads to a gradual but potentially fundamental reconfiguration of roles, often without immediate changes in job titles or employment levels. Therefore, it is increasingly vital for workers to understand how AI impacts key tasks in their roles, the complementary tasks that gain prominence, and the new necessary skills to enhance their individual productivity.
Preliminary Labor Market Signals Indicate AI’s Pressure on Entry-Level Roles
Aggregate employment data remains generally strong; however, preliminary evidence suggests a contraction in entry-level roles in AI-exposed occupations over the last several years. While the extent of AI’s direct role in the downturn remains debated, these positions may be particularly vulnerable to impact from AI because they involve routine and largely standardized tasks that AI may be able to perform. Whether this long-term impact on entry-level jobs is a bellwether to a broader impact to come is currently hard to ascertain. Nevertheless, the implications remain significant, in part because AI may disrupt the pipeline of talent and the well-established pathways for career progression in certain industries.
The Extent of AI Adoption Will Be One of the Principal Determinants of the Impact on Labor Market Outcomes
The potential impact of AI on the economy appears to depend on what the technology can do. However, one must not overlook the importance of adoption, which is neither pre-ordained nor even. Adoption varies across sectors and organizations, shaped by incentives, workflows, and institutional constraints. The balance between augmentation and automation will be determined by these implementation choices, and who participates in making them, rather than technological inevitability.
Shifting Skill Demand in the AI-Driven Economy
As AI has started to reshape the skills required in the economy, it is inevitably raising questions about the type of training and roles that may have more enduring value. In some AI-exposed occupations, new digital and technical skills, as well as durable human capabilities—such as critical thinking, leadership, and adaptability—may be key to prosperity. In other emerging and expanding industries—such as new AI-focused job categories and construction to support AI infrastructure—the supply of skilled labor is insufficient.
Traditional Educational Pathways and Trainings May Fall Short in Meeting the AI Moment
The current education, credentialing, and training systems remain largely built around front-loaded learning in the lead up to a degree and career, leaving them misaligned with rapidly changing skill demands. To keep pace, these systems must be reimagined to leverage AI as a training mechanism and to produce a balanced mix of foundational capabilities and adaptable domain-specific skills.
The Distribution of AI’s Gains Need Not Be Concentrated, and Can Be Shaped by Implementation Choices
AI can broaden access to high-productivity tools and raise worker performance, but it can also concentrate gains among firms, regions, and workers with early access and complementary skills. The outcome will depend on how AI is deployed, who benefits from productivity gains, and how quickly workers can transition skill sets and jobs.
The Need for Coordinated, Whole-of-Nation Action
No type of institution—government, industry, or educational—can manage the AI-powered transition alone. The speed, scale, and complexity of AI’s impact require coordinated action across sectors and actors to align incentives, invest in workforce development, and ensure broadly shared benefits. The window for proactive intervention is open.
Launching the Future: The AI Advantage Series on Coursera
In a direct effort to provide immediate and scalable training opportunities, we also offer a set of 5 new courses designed by SCSP, in partnership with Coursera. Developed in response to invaluable feedback to previous set of SCSP courses, this series is specifically designed to bridge the gap between foundational literacy and advanced strategic application.
These five additions bring SCSP’s total offerings to eight that we have launched with Coursera in the last year. These courses provide students and national security professionals with elite, self-paced AI knowledge. By balancing brevity with strategic depth, they ensure the American workforce is equipped to lead through rapid technological change.
The five new courses are short, modular and interactive, giving learners the maximum flexibility to chart their own AI journey.
Human-Machine Teaming for Government: An AI “boot camp” for current and aspiring public servants covering LLM fundamentals, ethical frameworks, and practical prompt engineering to automate routine tasks.
AI in Global Competition: The Defense Dimension: An analysis of how AI is reshaping the battlefield, from autonomous systems to the U.S.-China military-technological race.
AI in Global Competition: The Intelligence Edge: A deep dive into how AI and OSINT tools like Project Maven are revolutionizing the intelligence cycle and strategic warning.
The Future of Intelligence: Navigating AGI and ASI: A leadership-focused look at the trajectory of Agentic AI, Artificial General Intelligence (AGI), and Artificial Superintelligence (ASI).
Fusion Explained: The Geopolitics and Future of Energy: An exploration of the science of fusion and why it remains a critical national security imperative for clean, limitless power.
Alongside our standard programs, SCSP provides free access to the entire Coursera catalog. We encourage early enrollment as free slots are limited, allowing you to chart your own AI training journey.
Find the right course for you and sign up at https://www.scsp.ai/coursera-offerings/.
In case you missed it, The Washington Post Intelligence Briefing covered the AI and the Future of Work’s preliminary findings in its May 7 newsletter. You can read the coverage of the report, the AI + Expo and other topics related to AI.
Good to see this move from findings to a national-action framing. Most of the AI-and-work conversation stalls at diagnosis, so the shift matters. The reskilling pillar is where I'd push back. National plans almost always treat retraining as the release valve: displace at the entry level, retrain people into something new. But the capabilities that survive automation, judgment, relationships, knowing which problem is actually worth solving, don't come out of a six-month course. They accrue over years, in precisely the junior roles AI erodes first. So you break the ladder, then ask people to climb it. There's a second gap the national framing tends to miss. If AI makes one person five times more productive, the firm doesn't need five times the output, it needs fewer people. Those gains flow to capital, not to the worker who dutifully mastered the tool. A strategy that indexes on individual upskilling without touching aggregate demand for labour is answering the wrong question. Worth naming that explicitly before the policy hardens around it. You can master the tool that replaces you.