What does 2025 tell us about 2026?

Artificial intelligence is transforming how we work, demographics are reshaping the talent pool, and both the public and private sectors are rethinking how they use data to make smarter decisions. To help deliver clarity for the next move forward, we asked eight Lightcast thought leaders to look forward and make predictions for the new year.

Scroll through to see their expert perspectives—ranging from the emergence of Talent Intelligence as a critical business discipline, to a renewed focus on skills-based strategies, to the essential role of government data in shaping economic opportunity—or use the links on the left sidebar to jump ahead.

Focus: Demographics and People Migration

2025 ended up being a wash out year as early year promise was wiped out by tariff chaos. With everyone stalled, conspiracy theories abounded, the biggest being AI was stealing jobs when in fact, companies hesitant to hire, took their accumulated excessive profits from prior years and spent those monies on various investments including AI.  We see this in spiking investments in industrial machinery which are outside of the computer and electronics segment (which has only seen slow growth). Also, employment in semiconductors plunged to lows not seen until before AI kicked off in ’22. With this backdrop, what can we expect in 2026?

Let’s divide this into low-risk and high risk:

Low risk:  First half of the year, stability for businesses starts to return and tariffs find their holding ground. We would not expect much growth but do expect companies will be more able to see their future and act accordingly. Labor demand has been pent up with no real hiring for the past several years. It hasn’t been a recession because retirees leaving the labor force and deportations have lowered labor supply, otherwise unemployment would’ve likely spiked. However, when companies try to hire for non-degreed jobs, they will find almost no one looking for work. That said, nothing stays down forever but the US is going through a fundamental change in population dynamics which likely means milder GDP growth from here on out. But milder growth is the prediction despite some people predicting a recession.

High risk:  After a difficult first year setting policies using the “door-in-the-face” technique, policies will become friendlier in 2026. Immigration evolves and likely even returns in bits as we hit the mid-point of the year. Now feeling like the earth is no longer moving beneath their feet, various segments start to strengthen and hiring picks up as the year goes on.  Interest rates likely stay stable, so investments increasingly trickle away from AI into other industries that have been cash starved for 2 years which helps white collar hiring pick up. AI, which carried the GDP load in ’25 without adding any net jobs and incinerating billions of dollars, starts to fade with no clear line to products or profitability. Housing remains unaffordable in most of the country and wages will barely rise to help out. And now the riskiest statement of all. 2026 will be seen as an acceptable year. Not bad, not great. This will probably come back to haunt me.

Focus: AI Trends

2025 has been the year of the AI hype. Everyone has been talking about it - especially in the C-suite - while action on the ground has often been fragmented, experimental, and bottom-up. Workers, educators, and policymakers have been trying things out without clear organisation-level strategies to guide them.

This year the biggest question around AI has been: so what?? How do we use it to drive change and efficiency? How do we prepare for this change? And is this bringing value or is it just a new hype?

While I often say I don’t have a crystal ball, I think next year we will start getting answers to some of these questions. 

2026 predictions:

• From AI hype to clarity of action: If 2025 has been the year of front-line experimentation and innovation, 2026 will be the year where AI implementation will become more structured. Facing uncertain labour market conditions, companies will want to focus on what works and how to scale it. We will see company-wide initiatives to leverage how to embed AI in different job functions - what to prioritise and how to change it - and company-wide AI training will play a central role. I believe we will also see educators and the public sector taking organisation-wide approaches to AI, with clearer guidelines about AI use and AI ethics.

• A stronger link between education and business. One thing that emerged clearly in 2025 is that the scale of change - and noise - brought forward by AI require a different - and stronger - way of connecting education and work. From understanding what’s hype and what’s real, to figuring out how to adapt to these changes, we have seen AI pushing education and businesses to finally work together in 2025 - and I suspect they will get much closer next year. 

• Organizations will start walking their talk when it comes to skills. AI has accelerated a trend that’s been going on for years: the slow erosion of the value of degrees. On top of that, one of AI’s biggest labour-market effects this year has been the explosion in low-quality applications. AI-generated CVs are flooding recruitment, increasing costs and reducing hiring quality. Companies are bearing the brunt of this - but this has also implications for education and the public sector. Skills will finally emerge as a solution and all stakeholders in the labour market will start walking their talk about skills - from skills-based businesses to skills-based education and the link between the two. 

Not sure if this will happen in 2026 or if we’ll have to wait till 2027 - my last prediction is about AI moving from its own bubble to something that is connected to other labour market trends. People still talk about AI as if it’s the only thing in the labour market - forgetting the huge implications (and barriers) it has in connection to other macro trends, such as the environment, labour shortages and demographic shifts. It will only be when this happens that we will finally see what the transformational change AI is promised to have will actually look like in practice. 

Focus: Enterprise - US

Most predictions for a new year are a reframed synopsis and diagnosis of the present year. I will try not to do that. I usually try not to make predictions, but rather deductions based on the information we know now. 

First, let’s revisit some predictions from last year, and the implications they have for 2026:

• In 2025, while AI was predicted to threaten blue-collar jobs, it's actually the white-collar workforce that faces some displacement… [and] traditional large tech firms will continue their "efficiency" mandate… The government's efficiency work could lead to the job market being flooded white collar resumes, and it will take a while for the private sector to accommodate this influx of labor, putting pressure on wages and forcing job mobility. 

• Swift deregulation will lead the broader economy to grow in 2026, but will actually hurt the economy in 2025. 

• People analytics and workforce planning for most firms, though, will focus on efficiencies, productivity, and AI use cases. 

I think it would be hard to argue that 2025 didn’t see some of these predictions come true, and they have implications for 2026. The first implication is: From all the corporate layoffs and government efficiency work and potential AI impacts (debatable) on worker displacement, there is a glut of white collar talent on the market, and that is expected to continue into 2026. If you look at JOLTS data, since 2022 both hires and separations have been reducing in a linear fashion. Barring some exogenous shock to the economy, this trend will likely continue in 2026. However, with the pressure on white collar workers to continue to produce more while hiring is still down, pressure on employee burnout and worker dissatisfaction should rise in 2026.

Second, while the economy did struggle at times in 2025, many of the tariff-based commitments from other country’s governments and corporate leaders for more domestic investment in manufacturing and buying of US-based goods and services should materialize in 2026 – especially if cost of borrowing (i.e., interest rates) continue to decline, even modestly. I expect two things to happen as a consequence: 1) big investments will spark some economic growth in the US heartland, and 2) some of those “commitments” will fail to materialize into real action. The former will make site selection efforts and location strategy an important topic again for HR teams in 2026, and you may even see an increase in corporate relocations, for tax incentives, cost of labor, and political reasons.

What does this mean for HR, people analytics, workforce planning, and talent intelligence functions? 2026 will be the year of “AI workforce transformation”. Pay attention to that phrase. It will be the focus of HR in 2026. And I think you will begin to see the rise of the term “Work Intelligence” as the successor to many HR specialist terms I mentioned before. The “Work Intelligence” revolution will begin in 2026, and will likely be a multi-year project about reconstituting the HR function for how to add value with AI technology – while still being mindful of costs and producing value. Get ready for a new fun ride.

Focus: Education

2025 has been a disruptive year for higher education, but I predict that what we saw in 2025 is only a taste of what’s to come in 2026. Here are six of the ways I expect to see acceleration across nascent change vectors. 

1. Funding gets bigger — and more tied to outcomes

More states and the feds will put “outcomes strings” on new dollars (completion, earnings, equity), especially as Workforce Pell rolls out.

Trend anchors:

• State and local funding for higher ed grew 37% over five years, reaching $$139.1 billion. That’s roughly the size of the economies of Panama and Costa Rica put together. It will be unsustainable to dedicate those resources over time without clear ROI expectations. 

• By FY 2023, 28 states used performance-based funding, and those formulas now drive roughly 9.5% of four-year and 10.2% of two-year public funding nationwide—up from almost nothing a decade ago.

2. Affordability fights shift from loans to up-front aid design

As the policy fight moves from “how to repay” to “how to price,” we can expect tougher scrutiny of net price, better-targeted grants, and sharper value tests on aid programs.

Trend anchors:

• Though the maximum Pell Grant has risen about 28% over the past 10 years, its real value has barely budged after inflation.

• States and institutions now provide about $100B/year in grant aid, a scale large enough that redesigning those dollars is one of the most likely next levers to pull on affordability.

3. “AI-fluent” becomes the new baseline skill expectation

AI literacy will no longer be niche; it will be a default expectation in business, health, and public-sector roles, driving both curriculum refreshes and new credential types.

Trend anchors:

• Job postings requiring AI skills jumped 73% from 2023–24 and another 109% from 2024–25, according to Lightcast data.

• A Lightcast/Brookings analysis finds AI-related postings have grown at an average ~29% annually over 15 years, far outpacing overall job-posting growth.

These curves almost force institutions to move from “we offer an AI elective” to “AI is embedded in every major.”

4. Non-degree credentials become a parallel mainstream system

Non-degree credentials (certs, licenses, bootcamps, micro-credentials) will be treated less as “alternatives” and more as a parallel system woven into both pre- and post-bacc pathways.

Trend anchors:

• Annual attainment of vocational certificates and professional licenses both saw roughly a tripling of annual non-degree attainment rates over the past 10 years.

• New Pew data suggest up to one-third of U.S. adults now hold some kind of non-degree credential, from certificates to licenses and badges.

5. Apprenticeships and “earn-and-learn” degrees move from pilot to policy

Expect more apprenticeship-based degrees and structured “earn-and-learn” tracks, including options that stack into or sit alongside bachelor’s pathways.

Trend anchors:

• Active Registered Apprenticeships have climbed from about 360,000 in 2015 to over 667,000 in 2024, an 85%+ increase, with expansion into IT, healthcare, and business roles—not just the trades.

• States are now piloting apprenticeship-based degree models (e.g., Reach University) and embedding them in broader “pathways” initiatives.

6. Workforce Pell + alternative providers reshape the accreditation perimeter

The new Workforce Pell program will accelerate the entrance of non-traditional providers (bootcamps, employer academies, online platforms) into Title IV–eligible ecosystems, forcing states and accreditors to define quality at the program level.

Trend anchors:

• The “One Big Beautiful Bill” that created Workforce Pell represents a huge shift from aid tied to institutions to aid tethered to specific, short programs and external quality screens defined by the states. 
  

Summary:

Pulling back up, the throughline becomes clear: higher ed will be pushed to deliver value in smaller, stackable doses; to show that value in concrete ROI terms; and to align more directly with the evolving skill demands of the labor market.

Focus: Public Sector - US

For workforce and economic development leaders, 2026 is going to be the year of outcomes: outcomes-focused data, decisions, and programs. After a year of upheaval and limited funding, the federal, state, and philanthropic grants that emerge in 2026 will flow to organizations and agencies that prove ROI for learners, workers, and their communities and are aligned to local industry demand. This will be exhibited in the rollout of Workforce Pell, which mandates stringent accountability metrics to participating short-term or non-degreed credential programs. This is just one reason why building the right data infrastructure is so vital.

Related to this, we’ll also see more public sector leaders shed either/or thinking and strategies in 2026. The successful workforce training approach is not based exclusively on skills or degrees (it’s both or skills-first, not skills-only). It’s not about relying on one type of data vs. another type to make the best decision (it’s about triangulating multiple data sources). It’s not just about attracting data centers (many regions will lose out if they focus too much on the AI buildout). It’s not relying only on public or private funding (where possible, it’s both).

Goodbye, either/or thinking.
Hello, both/and strategy and action centered on outcomes.

Focus: Government data

In 2025, private businesses, policymakers, and individuals voiced many complaints about the quality and timeliness of official US government data. But when the government shutdown temporarily and stopped producing key datasets, its absence was immediately felt. The lack of the government’s “gold standard” benchmarks created “foggy” blind spots, especially during such a critical turning point in the economy, at a time when the Federal Reserve needed every piece of information it could get.

Analysts poured into private sources of data but soon realized some of the shortcomings. For example:

• Many private datasets have coverage gaps, often overrepresenting large firms, urban areas, or higher-income workers while missing small businesses, rural regions, and underrepresented populations.

• Private data are subject to methodological changes without warning or transparency (shifting customer bases, algorithm updates, vendor changes), which can introduce instability, breaks in trends, and make historical comparisons difficult.

• Private data are shaped by commercial incentives, meaning what gets measured is often what is profitable or easy to collect, not necessarily what is most economically meaningful.
  

Yes, public data aren’t perfect either. But they are provided as a public good, without a lens toward profit or to influence any particular outcome. They are objective, wide in breadth, and aim to cover most corners of the economy–rural areas, small businesses, and underrepresented populations. They are meant to provide both short-term signal and historical context for identifying structural changes in the broader economy.

Prediction: People will continue to question, critique, and scrutinize government data (and rightly so), but they will more clearly recognize its indispensable role. They will increasingly respect it for what it is, what it is intended to do, and the critical part it plays in providing reliable economic information.

Focus: Skills & The Future of Work

My big prediction for 2026 is that leaders will no longer be asking whether AI is changing work; they’ll be asking why their operating model still assumes it hasn’t. By next year, the pace of change will make anything else untenable. The organisations that thrive will be the ones that stop treating AI as an experiment and start redesigning work with intention, clarity, and honesty.

To understand the shift underway, we have to start with the reality of where AI is hitting. Despite the comforting narrative that “AI won’t take jobs, it will only take tasks,” the truth is more uncomfortable. In several fields, AI is already eliminating jobs outright. The logic is simple: AI automates tasks at scale, and when enough tasks within a role disappear, the role inevitably collapses. If organisations cannot see work at the task level, they cannot see what AI is really changing.

Yet even as tasks have quickly entered the conversation, the foundation of the talent ecosystem remains unchanged. Skills still power every major HR technology workflow - matching, mobility, development, hiring and workforce planning. Skills are still the operating system of people decisions, and we are not replacing that architecture overnight. The lesson from the past two years is clear: stabilise your foundations, align to business outcomes, and stay pragmatic.

Which is why the real breakthrough for 2026 will not come from choosing between skills or tasks, but from connecting them. Organisations will need to understand what people can do (skills), what they actually do (tasks), and how both shift as AI rewrites work in real time. When those elements come together, workforce strategy finally becomes inseparable from business strategy, and that is the operating model reset that will define 2026.

Focus: Talent Intelligence as a Discipline

As we enter 2026, talent intelligence finds itself caught in a defining paradox: organizations demand unprecedented efficiency and scale while conducting layoffs and working with technology that simply isn't delivering. With 69% of talent acquisition leaders reporting their GenAI investments haven't produced expected gains (as seen in a recent Talent Intelligence Collective mini poll), despite $49 billion flowing into GenAI startups in just the first half of 2025, the gap between promise and reality has never been starker. But something fundamental is shifting beneath this surface tension. The flat labor market forcing efficiency demands is simultaneously driving a strategic pivot that will reshape our profession. Organizations can no longer hire their way to new capabilities—they must build them internally. This reality is breaking talent intelligence free from its acquisition silo and forcing genuine integration with workforce planning, L&D, and strategic operations.

Talent intelligence is evolving from a specialized function into the connective tissue binding all people operations together. The numbers tell the story: in our recent work Beyond the Buzz we saw that while HR shows just 2% of job postings requiring AI skills today, it's experiencing 66% annual growth. The 65% surge in People Analytics capabilities and 81% jump in Performance Management skills signal organizations recognizing they must develop talent, not just source it. This isn't merely collaboration—it's fundamental convergence. The lines between talent intelligence, workforce planning, and capability development are dissolving. When you can't recruit your way to new skills, intelligence must inform development strategies. When markets shift weekly and GenAI transforms industries in months, planning can't remain a quarterly exercise owned by central teams.

We're witnessing the radical decentralization of workforce planning and talent intelligence capabilities. The traditional model—centralized teams conducting analysis and distributing insights—becomes obsolete in a BANI world (Brittle, Anxious, Nonlinear, Incomprehensible) where competitive advantage evaporates in weeks. Business leaders need agile intelligence embedded directly in their teams, making decisions in real-time. This drives the emergence of embedded planning cells within business units, armed with centralized tools and data but empowered for rapid, contextual decisions. Talent intelligence professionals are becoming enablers rather than gatekeepers, building capabilities throughout the organization rather than hoarding expertise in central functions. The question shifts from "What insights can we provide?" to "How do we enable intelligent decision-making at scale?"

This decentralization strategy succeeds or fails on one critical factor: data foundations. You cannot safely distribute AI-powered intelligence tools without rock-solid infrastructure—consistent taxonomies, validated skills frameworks, and robust governance structures. The push to embed capabilities throughout organizations demands data clarity at an unprecedented level. Organizations are finally acknowledging that internal talent data means nothing without external context, that isolated skills taxonomies become dangerous blind spots.  While AI capabilities explode across every function, AI Ethics and Governance skills appear in less than 1% of postings. Organizations are building powerful decentralized capabilities on foundations of sand. The need isn't just for clean data—it's for controlled, consistent, scalable foundations that enable decentralized execution without descending into ungoverned chaos.

The Path Forward

The year ahead will determine whether talent intelligence matures into the strategic function it's always promised to be. Success demands we move beyond siloed expertise and quarterly reports. It requires building data foundations robust enough to support distributed decision-making, creating frameworks that enable rather than constrain, and establishing governance that scales without suffocating agility.

Those who navigate this transformation—evolving from insight providers to capability builders, from centralized experts to embedded enablers—won't just survive the disruption ahead. They'll define what talent intelligence becomes.