AI and the Future of Work: What’s Really Happening to Jobs in 2026

Introduction

The question dominating the 2026 World Economic Forum at Davos was not whether AI would change work — that debate is settled. It was how fast, how profoundly, and what kind of society emerges on the other side of the transformation.

IMF Managing Director Kristalina Georgieva set the tone when she described AI as “a tsunami hitting the labour market.” Her warning reflected an institution no longer hedging — the IMF’s own data now shows 40% of global jobs and 60% in advanced economies are exposed to AI disruption, with entry-level roles bearing the heaviest impact.

The data arriving in early 2026 is more complex than either the optimists or the pessimists predicted. The labor market is experiencing something genuinely new: simultaneous acute shortage of certain skills and abundant displacement in others, within the same companies and often the same departments.

One in 10 job postings in advanced economies now requires at least one new AI-related skill. Jobs requiring those skills carry a 56% wage premium over comparable roles, according to PwC’s 2025 Global AI Jobs Barometer — up from 25% just a year earlier. Industry estimates put unfilled AI positions at over 1.6 million globally, with a demand-to-supply ratio of 3.2:1. And yet: companies are simultaneously laying off entry-level workers, describing how AI will let them “do more with less.” Half of employers plan to reorient their business in response to AI. Fully 41% anticipate reducing their workforce in areas where AI can automate tasks, per the WEF’s Future of Jobs Report 2025.

Both things are simultaneously true. The question is which effect dominates in aggregate — and for whom.

What the Economists Are Finding

The macroeconomic evidence on AI and employment is still early — the technology has not been deployed at scale for long enough to see full labor market effects — but several major findings have landed.

The WEF’s net job projections are now concrete. The Future of Jobs Report 2025, which surveyed over 1,000 employers representing 14 million workers across 55 economies, projects that by 2030 AI and related trends will displace 92 million jobs while creating 170 million new roles — a net gain of 78 million positions. But that net figure masks enormous churn underneath: entire categories of work are disappearing while new ones emerge, and the skills required for the new roles bear little resemblance to those they replace. The report found that 40% of current workforce skills will become obsolete within five years.

At Davos 2026 itself, WEF released a separate analysis — “Four Futures for Jobs in the New Economy: AI and Talent in 2030” — modeling four scenarios ranging from “Supercharged Progress” to an “Age of Displacement.” Which scenario materializes depends on policy choices being made now.

The IMF’s 2026 assessment confirmed that AI is reshaping labor markets primarily through task automation rather than wholesale job elimination. Rather than wiping out job categories entirely, AI is automating specific tasks within jobs, changing what workers do rather than whether they have jobs. This is consistent with historical technology adoption patterns — but the speed is not. Of workers in AI-exposed roles, roughly half could benefit from augmentation; for the rest, key tasks are being automated, leading to lower wages and slower hiring.

Goldman Sachs’ landmark 2023 analysis (still the most comprehensive published estimate of its kind) projected that generative AI could automate roughly 25% of work tasks across the US economy and 24% across Europe, with significant variation by occupation. The report estimated 300 million full-time jobs globally are exposed to some degree of automation, and that AI could raise global GDP by 7% over a decade. Goldman noted that past waves of automation created more jobs than they displaced — but acknowledged that the pace and breadth of AI’s capabilities are without historical precedent.

PwC’s 2025 AI Jobs Barometer — analyzing roughly one billion job advertisements across six continents — found that AI-exposed sectors are seeing nearly four times faster productivity growth than the broader economy (27% growth from 2018 to 2024, compared with 7% from 2018 to 2022 in those same sectors before AI penetration). The skills employers seek are changing 66% faster in AI-exposed occupations than elsewhere. Jobs in highly AI-exposed industries are still growing, but the composition of those jobs is shifting rapidly.

---

The Layoffs Nobody Is Talking About: Entry-Level Roles

The most concerning near-term employment dynamic is the hollowing-out of entry-level professional roles — the positions that have historically served as the on-ramps into careers.

The IMF’s January 2026 data puts numbers on what many suspected: automation is two to three times more likely to affect entry-level positions than their managerial counterparts. Marketing analysts, sales representatives, and graphic designers are being hit harder than their senior equivalents. Challenger, Gray & Christmas data shows AI cited as a factor in roughly 55,000 US job cuts in 2025 alone.

In knowledge work sectors — software development, marketing, legal, finance, consulting — the tasks that AI handles most effectively are precisely those that junior employees have historically done:

• Junior developers were hired to write boilerplate code, fix simple bugs, and write documentation. GitHub Copilot and similar tools now do these tasks in seconds.
• Marketing coordinators drafted social media posts, wrote product descriptions, and researched competitors. AI writes these faster and cheaper.
• Junior lawyers reviewed discovery documents, researched precedents, and drafted first versions of contracts. Harvey and similar legal AI tools do this at scale.
• Financial analysts aggregated data, built first-version models, and wrote preliminary reports. AI produces these in minutes.

The consequence is appearing in hiring data: many companies are hiring fewer entry-level positions, expecting existing staff augmented by AI to absorb junior-level work. Tech companies that laid off tens of thousands in 2023-2024 have been selectively hiring in 2025-2026, but new hires are concentrated at senior levels and in AI-specific roles — not restoring the junior workforce that was cut.

This is not a temporary dislocation. It is a structural change in how knowledge work careers begin, with profound implications for social mobility, skills development pipelines, and how the next generation of professionals accumulates expertise.

---

The New Hot Roles: What AI Has Created

For all the displacement, AI is creating roles that did not exist three years ago and generating employment growth in several areas.

AI/ML Engineers and Researchers. The people who build AI systems remain in extraordinary demand. Total compensation at frontier AI labs for PhD-level researchers now reaches $500,000 or more — including base salary, equity, and bonuses. H-1B visa filings reveal that OpenAI technical specialists can earn up to $530,000 per year, while Anthropic research engineers reach $690,000. OpenAI has offered retention bonuses as high as $1.5 million. Even more modest AI engineering roles command significant premiums, and the talent wars have now reached intern-level hiring.

Prompt Engineers and AI Product Managers. The interfaces between AI capability and business application — people who understand both what AI systems can do and what specific business problems need solving. Still a nascent profession, with significant variation in what the role actually involves, but increasingly recognized as critical.

AI Trainers and Quality Specialists. The humans who label training data, evaluate model outputs, write RLHF (Reinforcement Learning from Human Feedback) preference data, and assess model safety. Companies like Scale AI (valued at over $20 billion), Appen (with a global workforce exceeding one million contributors), and Surge AI (which has contracted over 20,000 professionals with doctoral degrees) employ large teams of knowledge workers performing these functions. The data labeling sector alone is projected to reach $1.89 billion in 2025, growing at 23.6% annually.

AI Safety and Ethics Professionals. Companies building and deploying AI are under growing pressure from regulators, customers, and their own employees to demonstrate responsible AI practices. AI safety researchers, responsible AI teams, and ethics specialists are being hired across the industry.

Automation Engineers. As organizations automate workflows using AI, they need engineers who can design, build, and maintain the automation — integrating AI tools into existing business processes and IT systems.

AI Healthcare Specialists. Radiologists who interpret AI-assisted diagnoses (85% of whom now believe AI will improve consistency and patient outcomes), clinicians who validate AI triage recommendations, and pharmaceutical scientists who evaluate AI-generated drug candidates. The radiologist’s role is evolving into what some describe as a “diagnostic orchestrator” — and the profession is growing, not shrinking, as demand increases for professionals who can work effectively alongside AI.

---

The Skills Reskilling Imperative: What Workers Must Do

For workers navigating this transformation, the evidence suggests several practical orientations.

Learn to work with AI, not against it. The workers who are thriving are those who have integrated AI tools into their workflows and use them as force multipliers — producing three to five times the output of colleagues who haven’t adopted AI tools. PwC’s data shows this productivity differential is real and measurable at the sector level, and it is translating into career outcomes and wages.

Develop skills that AI genuinely cannot replace. Complex interpersonal skills, creative judgment, contextual wisdom, ethical reasoning, physical presence and dexterity — these remain distinctly human. The worker who combines AI augmentation with irreplaceable human capabilities has the most defensible position.

Treat AI literacy as a baseline professional skill. Understanding how AI systems work well enough to use them effectively, evaluate their outputs critically, and identify where they are likely to fail is increasingly a baseline expectation across professional roles. The IMF found that AI literacy carries a labor market premium across virtually every sector — job postings requiring four or more new AI-related skills command wage premiums of up to 15% in the UK and 8.5% in the US.

Specialize in high-value domains. AI is commoditizing general competence in many fields. Specialized, deep expertise — the cardiologist who understands cardiac physiology beyond what AI provides, the lawyer who knows the specific regulatory environment of a particular industry — retains premium value precisely because AI makes generic competence abundant.

Build network and interpersonal skills. As AI handles more of the analytical and writing work that traditionally justified professional relationships, humans who can build genuine trust, navigate complex organizational dynamics, and lead effective teams become more valuable, not less.

---

The Geographic Inequality of the AI Transition

The AI employment transition is not uniform across geographies. It is concentrated in ways that could widen existing inequalities.

High-income, English-speaking labor markets face the most immediate disruption, where knowledge work is most prevalent and AI tools — primarily trained on English data and optimized for Anglo-American professional contexts — are most effective.

The country-level readiness gap is stark. The IMF’s workforce preparedness index ranks Finland, Ireland, and Denmark highest. At the other end, Germany could see 70% of its AI jobs unfilled by 2027. India’s AI sector may have 2.3 million openings by 2027, with only 1.2 million professionals in the talent pool. The WEF found that regions with high AI-skill demand are already seeing employment in AI-vulnerable occupations fall 3.6% lower after five years compared to regions with lower AI demand.

Young workers bear a disproportionate burden. The tasks most amenable to AI automation are disproportionately performed by younger, less experienced workers — the same workers who depend on those tasks to build experience and advance their careers.

Developing economies face a complex picture. AI tools that work best in English may not displace workers in other languages as rapidly. But business process outsourcing — call centers, data processing, software development — that provides entry-level employment for educated young people in India, the Philippines, Eastern Europe, and elsewhere is highly automatable. The IMF estimates that developing economies are more exposed to AI-driven disruption in their tradeable service sectors than they are positioned to benefit from AI-augmented productivity.

---

What Employers Must Do

Organizations managing this transition have responsibilities that go beyond pure efficiency optimization. The WEF’s data provides a framework: 77% of employers plan to upskill workers, 70% expect to hire staff with new skills, and nearly half plan to transition staff from declining to growing roles. But aspiration and execution are not the same thing.

Invest in reskilling — seriously. Companies that displace workers through automation and invest in reskilling those same workers demonstrate both ethical commitment and enlightened self-interest. IBM has committed to skilling 30 million people globally by 2030 and has already reskilled 30% of its own workforce in AI competencies. Amazon invested $1.2 billion through its Upskilling 2025 program and expanded to “AI Ready,” offering free AI training to two million people. Microsoft committed $4 billion over five years through its Elevate initiative, targeting 20 million people earning AI credentials.

Redesign jobs, don’t just add AI. The organizations getting the most value from AI are not those that have bolted AI tools onto existing job descriptions but those that have thoughtfully redesigned workflows — determining what humans should do, what AI should do, and how they collaborate.

Maintain hiring pipelines. Using AI efficiency as a reason to freeze hiring across the board, rather than selectively where AI genuinely replaces functions, creates organizational brittleness. The WEF’s Davos 2026 white paper found that 54% of executives anticipate AI will displace existing roles, but only 24% expect net new job creation and just 12% foresee higher wages. Organizations need a continuing pipeline of new talent even as the composition of that talent shifts.

Be honest with workers. The most corrosive dynamic in AI workplace transitions is management claiming AI will “augment” workers while privately planning to reduce headcount. The WEF data shows 40% of employees express concern about AI-driven job loss. Workers who know the plan can prepare; those who are misled cannot.

---

Conclusion

The AI employment transition of 2026 is not the apocalyptic job destruction that some predicted — nor the painless enhancement that AI optimists promised. It is a genuine structural transformation of knowledge work, with 92 million roles projected to disappear and 170 million new ones emerging by 2030. The net gain of 78 million jobs is real, but so is the pain of transition for those whose skills become obsolete.

The costs are falling disproportionately on entry-level workers, younger demographics, and developing economies. The benefits are flowing to those with AI skills (who command a 56% wage premium), specialized domain expertise, and the adaptability to continuously learn. The 40% of workforce skills projected to become obsolete within five years is not a distant warning — it is a description of what is already underway.

The policy and business choices made in the next three to five years will determine how the benefits and costs are distributed. Those choices are not predetermined by the technology. They are being made now, in hiring decisions, training investments, regulatory frameworks, and social policy — by companies, governments, and individuals navigating the same unprecedented transformation.

---

Sources & Further Reading

• New Skills and AI Are Reshaping the Future of Work — IMF (Jan 2026)
• IMF Staff Discussion Note: Bridging Skill Gaps for the Future (SDN/2026/001)
• WEF Future of Jobs Report 2025
• WEF: Four Futures for Jobs in the New Economy — AI and Talent in 2030
• Jobs and Skills Transformation: What to Know at Davos 2026 — WEF
• PwC 2025 Global AI Jobs Barometer
• Goldman Sachs: Generative AI Could Raise Global GDP by 7%
• IMF Chief Warns of AI ‘Tsunami’ for Entry-Level Jobs — Fortune (Jan 2026)
• 41% of Companies Plan to Reduce Workforces by 2030 — CNN (Jan 2025)