Part 3: Policy, Regulation & Legislation

Bridging the Skills Gap: A Defining Policy Challenge of Our Time

Shawn Olds

I took my first Artificial Intelligence (AI) course in 1992 while a Cadet at West Point. Over three decades ago, AI was still rising on the horizon. With the launch of Generative AI in 2022, AI is no longer on the horizon; it is actively reshaping the world of work. From automating routine tasks to augmenting creative roles, AI is disrupting industries at a pace that outstrips any past technological shift. The World Economic Forum estimates that 25% of jobs will be disrupted within five years, and up to 85% of workers could see a portion of their tasks impacted by Generative AI.¹

The question is not whether the workforce will change; it is whether our policies are prepared to guide that change.

Historically, moments of industrial transformation have widened inequality before closing it. This time, we have a chance to flip the script. With smart, coordinated governance, AI adoption can be paired with policies that elevate, not exclude, workers. But to do so, reskilling and upskilling must move from the sidelines to the center of our AI roadmaps.

This chapter frames workforce transformation as more easily and swiftly accomplished by public policy, one that requires cross-sector collaboration, forward-looking incentives, and a clear commitment to building for the future. By weaving human-centered principles into governance frameworks, we can create an AI future where progress is measured not just by productivity, but by participation. This chapter will look at why a policy approach can provide an efficient and effective path forward, it will provide recommendations on how that policy can be developed, and it will conclude with three use cases, both domestically and internationally, where policy has played an effective role.

The list of new challenges AI presents grows daily. Some of the challenges include, but are not limited to:

Twenty-three percent of current jobs will change by 2027.²
An estimated 12 million U.S. workers may need to change occupations by 2030.³
The Brookings Institution finds that over 30% of workers could see at least half of their tasks affected by generative AI. And up to 85% of workers could see at least 10% of their functions impacted.⁴

One of the biggest challenges is that most current statistics, like the above, are based on “today’s” AI, not “tomorrow’s” AI, and the technology continues to evolve at an unprecedented pace. Hence, these numbers are likely to grow significantly over the next 5 years.

History demonstrates that technology shifts can displace workers (i.e., printing press, automated loom, assembly line), but these shifts have also presented opportunities and evolutionary leaps for humanity. Some of the opportunities that we are seeing include:

AI could boost global GDP by 7%.
There will be an increase in the ability to upskill third-world workers, as well as workers whose primary job functions can be replaced by AI. African economies could unlock up to $100 billion in annual economic value across multiple sectors from gen AI alone.⁵
AI could add $3 trillion to the U.S. economy over the next decade.⁶
Rather than cutting jobs, AI can increase demand for fresh graduates. At Cognizant, a provider of information technology, consulting, and business process services, AI boosted productivity among junior developers by 37%, unlocking new roles in supervising and managing AI systems.⁷
AI is spawning new professions, from AI trainers and ethics consultants to system auditors, that did not exist a decade ago, highlighting this as a major human-machine collaboration opportunity.⁸

With some of the more recent technological advances, the internet, cell phones, etc., society could adapt over time. AI is moving at such a pace that workforce reskilling, upskilling, and education cannot be an afterthought. Through an appropriate public policy, we can fully capitalize on this next evolutionary jump.

The good news is that the American workforce in recent decades has embraced occupational change. My father had two jobs in his lifetime, and I was onto my third job by the time I was 28. The COVID-19 pandemic further illustrated this as it was a period of accelerated job and occupational switching not only in the United States but around the world. As companies closed, people were forced to look at new opportunities.

Recent research has also shown that the most impacted jobs are those of the youngest workers. Given the limited time they have invested in their careers, they tend to be the most flexible in finding new opportunities. Older workers, though, with more monotony and repetition in their jobs that AI can replace, may have a tougher time and may need resources to help them make that transition if they are to remain productive members of the workforce.

Reskilling as Public Policy for Responsible AI Adoption

Why policy? The scale and speed of AI disruption demand coordination and resources beyond what individual firms or workers can muster. Lifelong learning infrastructure, labor market incentives, and social support systems are public goods that will benefit from government coordination. Most leaders would prefer an employed workforce that pays taxes versus an unemployed population that they must support through social welfare programs.

1. Ensure Inclusive Adoption

Without intervention, AI could exacerbate social divides where workers with AI skills will thrive, but others may be left behind. Reskilling policies serve as a social equalizer ensuring everyone, from factory workers to finance professionals, can upgrade their skills for new roles. This is not a new concept; almost every U.S. president over the past four decades has campaigned on upskilling the workforce to ensure the country remains competitive.

When trade globalization threatened jobs, governments responded with Trade Adjustment Assistance programs. Today, a parallel “AI Adjustment Assistance” is being discussed to support those displaced by automation. Public policy can incentivize companies to accompany AI adoption with worker retraining.

2. Governance at Scale

To achieve responsible AI at scale, government action can align stakeholders and fill gaps. This can include funding training programs, setting skill standards, and using regulatory levers like workforce impact assessments or training requirements in AI procurement. Rapid AI advances raise questions about how employees will gain new skills to succeed in the future economy. A policy framework is needed to coordinate answers across the ecosystem with education systems, employers, labor agencies, and communities working in concert. Making workforce readiness an explicit goal of national AI strategies ensures that progress in AI innovation is matched by progress in human capability.

3. Economic Competitiveness and Innovation

Countries leading in AI are those that cultivate talent broadly. National policies for AI skilling are emerging as a competitive strategy. The U.S. and other countries have recognized that an AI-ready workforce is essential for innovation, cybersecurity, and economic security. Public investments in AI education and training yield returns and high productivity in new industries. Conversely, failing to prepare workers could slow AI adoption.

Key Policy Recommendations for an AI-Ready Workforce

1. Launch Public-Private Reskilling Initiatives at Scale

Governments can convene cross-sector partnerships to offer AI education and training to workers who will need it. Public-private initiatives leverage the strengths of each sector:

Government funding and coordination
Industry technology and expertise
Academic instructional capacity

The World Economic Forum’s “Reskilling Revolution” exemplifies this approach globally, engaging 370+ companies and 17 governments to empower 1 billion people with new skills by 2030, and already reaching over 680 million through committed programs.⁹ Nationally, policymakers can establish “AI Skills Councils” that include employers, universities, community colleges, and labor groups to align training curricula with market needs. It will be important to create easy access by ensuring online access as well as the ability to consume the material on a part-time basis, so lower-income households can continue to earn a paycheck while upskilling. IBM provides another example of this through its partnerships with NGOs. In September 2023, IBM pledged to train 2 million people in AI by 2026 with a focus on underrepresented communities through the free IBM SkillsBuild platform, learners worldwide access courses in AI fundamentals, chatbots, and even generative AI, earning digital credentials upon completion. This is part of IBM’s broader initiative to skill 30 million people by 2030. Since 2021 over 7 million learners have engaged with IBM’s courses²⁰.

2. Provide incentives for employer-led training

To encourage businesses to invest in AI skills, policymakers should offer tax incentives for training expenditures or hiring apprentice trainees in AI-related roles. Policy can encourage more firms to transition workers into new AI-augmented roles by lowering the cost of training through credits, grants, and recognition programs. Some firms will be incentivized on their own as PwC was. In mid-2023, PricewaterhouseCoopers U.S. announced a bold $1 billion investment to upskill all 65,000 of its employees in AI over three years. This firm-wide program is fully employer-driven, aimed at making PwC’s workforce fluent in generative AI tools and responsible AI practices²¹.

3. Strengthen the education to career pipeline with an AI Curriculum

While we should act on workforce goals now, acting strategically will make sure the pipeline of workers entering the workforce will not need these short-term initiatives. Educational policy should produce AI-ready graduates at all levels. This will require updating K-12, vocational, and higher education curricula to include AI literacy, data science, business intelligence, analytics, and critical thinking about technology. Given how easy it is for students to use GenAI to complete their work, educators have to rethink how they teach. Historically, we have always graded students on a product (an exam, paper, project); now, educators must grade on the process. The same technology that is forcing them to revamp the way they teach can help them ensure their students learn the fundamentals of their subject and then use AI to amplify those fundamentals. Combined with the critical learning and critical thinking skills they develop through the education process.

Another lever is offering student incentives, like AI scholarships or loan forgiveness for those gaining AI competencies in exchange for public service. By weaving AI training into the fabric of the education system, governments can ensure a continuous pipeline of talent.

Steps in this direction were taken in July of 2025 when OpenAI, Microsoft, and Anthropic announced an initial $23M investment on their part to create something called the National Academy for AI Instruction with the goal of training 400,000 K-12 educators, in practical AI skills by 2030. That is almost 10% of America’s entire teaching workforce.²²

4. Enable data-driven workforce planning and accountability

As society has more access to data, policymakers need to use it to drive their decisions and help them adjust in a fast-moving evolution of technology. This will mean investing in labor market intelligence systems that use AI to identify emerging skill needs and at-risk jobs in real time.¹⁰ By anticipating trends (e.g., noticing a surge in demand for AI ethics specialists, or a decline in certain routine roles), training programs can be updated proactively. The creation of public-private data partnerships, as recommended by the U.S. National Academies, would allow the sharing of data on skills demand, wages, and training outcomes to inform policy.¹¹ More importantly, it allows the government to understand when they can cease their initiatives and allow the system to run its course versus funding outdated programs.

Taken together, these recommendations provide short-term, medium-term, and long-term policy approaches for incentivizing learning, mandating responsibility, supporting transitions, building partnerships, and guiding with data. There are many possibilities of other initiatives that will be helpful, but these four form a foundation to build upon.

While many proposals are theoretical, given how new the technology is and how rapidly it is evolving, there are examples from around the world that we can replicate, adopt, or modify as appropriate, rather than needing to reinvent the wheel.

THREE CASE STUDIES

1. South Carolina's Apprenticeship Renaissance

South Carolina's success in expanding apprenticeships showcases effective ecosystem coordination involving state government, industry, and the technical college system. In 2007, the state launched “Apprenticeship Carolina,” a program with the SC Technical College System, to spur employer-driven training. Over the next decade, the number of registered apprentices in South Carolina grew from approximately 777 to over 8,000 annually.

Policy Measures

A key policy was the state apprenticeship tax credit of $1,000 per apprentice hired, available to any business with a registered program.¹² The simple incentive, coupled with free consultative support to help employers set up apprenticeships, lowered barriers for companies (especially small businesses) to participate. The state also leveraged federal grants to focus on high-growth fields, including IT and advanced manufacturing.

Results

South Carolina now leads the U.S. in the growth of apprenticeships, including programs in software development, cybersecurity, and Industrial AI. Industries that traditionally did not use apprentices, like professional services, have come on board. Employers benefit from a pipeline of loyal, trained workers, while individuals earn wages (and pay taxes) as they learn new skills. The model’s success has influenced national policy discussions, with experts citing South Carolina's credit as a model for federal incentives to encourage AI-aligned apprenticeships.¹³ By coordinating state agencies, colleges, and employers under a unified program, South Carolina created an ecosystem where everyone wins.

2. Finland's 1% AI Education National Strategy

In Finland, a unique collaboration between government, academia (University of Helsinki), and industry (tech firm Reaktor) led to the creation of the Elements of AI online course. Initially launched in 2018, its goal was to train at least 1% of Finns (55,000 people) in the basics of AI. The initiative was part of Finland’s broader aim to reposition its economy in the AI era following Nokia’s decline, focusing on broad-based AI literacy as a competitive differentiator.¹⁴

Approach

The Elements of AI course was offered free to the public, with content designed to be accessible to non-engineers (e.g., a dentist is profiled among successful learners).¹⁵ The Finnish government embraced the project, integrating it into the national AI strategy and actively promoting it to citizens as a civic learning challenge. Private companies provided sponsorship and helped with platform and content expertise. The course covers fundamentals of machine learning, neural networks, and societal implications, without heavy math prerequisites. As the then Minister of Economic Affairs, Mika Lintilä, noted, “We’ll never have as much money to be the leader in AI technology, but how we use it, that’s something different”, emphasizing human capital as Finland’s strength.¹⁶

Results

The initiative wildly surpassed its original goal. Within a couple of years, over 250,000 Finns (roughly 5% of the population) enrolled. Buoyed by this success, Finland offered the course to the European Union as a gift, aiming to train 1% of all EU citizens in AI. To date, the free course (available in multiple languages) has been taken by over 1 million people worldwide, from students to retirees.¹⁷Finland’s workforce has thereby gained a foundational awareness of AI, improving receptivity to AI adoption in workplaces and public services. Companies in Finland report that employees who took the course are more proactive in identifying AI use-cases in their jobs, indicating a cultural shift where AI is seen as an opportunity, not a threat. Finland’s case demonstrates the power of government-university-industry collaboration in upskilling a population at scale. Key success factors were: a clear vision, an easy-to-access learning platform, and strong public advocacy.

3. Singapore's AI skills ecosystem

Singapore offers a compelling example of a multi-pronged, executive-to-citizen training ecosystem driven by the government. Under its national AI strategy, agencies like the Infocomm Media Development Authority (IMDA) and AI Singapore (a government-funded R&D organization) partnered with tech giants (Microsoft, Intel) and local universities. Their mission was to prepare Singapore’s workforce and leadership for AI across all levels. In 2018, AI Singapore launched two flagship programs: AI for Industry (AI4I) and AI for Everyone (AI4E).¹⁸ AI4I was a subsidized three-month training for 2,000 tech professionals and company executives, teaching practical AI skills (like using Python for basic AI applications) to enhance productivity in their firms. Participants who completed the training earned an industry-recognized “Foundations in AI” certificate, and companies benefited as employees returned with actionable AI project ideas. AI4E, on the other hand, was a broad-based AI literacy drive targeting 10,000 Singaporeans from secondary students to working adults, introducing them to what AI is and how it can improve daily life and work.¹⁹

The content highlighted real-world use cases to make AI relatable and instill a mindset of innovation. Both programs were heavily supported by the government (with course fees largely subsidized by IMDA) and by industry partners providing curriculum and tools.

Results

Over a few years, these programs helped cultivate an AI-aware culture among Singapore’s workforce. Many small and medium enterprises (SMEs) sent their managers to AI4I, seeding AI pilot projects in sectors like logistics and healthcare that traditionally lag in tech adoption. Government agencies themselves enrolled public officers in new AI courses, aligning the civil service with the national AI agenda. For the general public, AI4E demystified AI and reached thousands of students, some of whom pursued further tech studies as a result. The city-state’s workforce is now frequently cited in global rankings as among the most digitally ready. By proactively upskilling, Singapore manages to implement AI solutions (e.g., smart city projects) with a workforce that can operate and maintain them, illustrating how policy-guided training underpins successful AI integration.

A Call to Action for Human-Centered AI Governance

AI is rewriting the rules of the labor market. But whether it becomes a force for economic mobility or deepened inequality depends on the choices we make now.

This chapter has outlined how workforce transformation is an economic necessity that can be empowered and catalyzed by appropriate policies. To equip society and the workforce for AI, public policy must step in where markets alone cannot: coordinating ecosystems and catalyzing change at scale.

The case studies from South Carolina, Finland, and Singapore show that we do not need to start from scratch, across geographies and political systems.

This is not about protecting old jobs; it is about preparing people for better ones.

Policymakers, educators, and business leaders now have a choice: treat AI as a wave to brace for, or a tool to shape a more just and capable society. Let us choose the latter. Let us choose humanity by investing in it.

Policy Lever

Objective

Approach

Example or Case Study

1. Public-

Private Reskilling

Initiatives

Scale accessible

AI training across

sectors

Cross-sector alliances: government funding,

industry expertise,