# AI Replacing Jobs: 30+ Statistics You Need to Know in 2026 > The most comprehensive collection of AI job replacement statistics for 2026. Data from Goldman Sachs, Oxford, McKinsey, WEF, and BLS - all in one place with expert analysis. ## Quick answer: How many jobs will AI replace? According to Goldman Sachs, approximately 300 million jobs globally are exposed to AI automation. However, 'exposed' doesn't mean 'replaced' - the OECD estimates 14% of jobs face high automation risk, while 32% will see significant task changes. The World Economic Forum projects 97 million new roles will emerge to offset 85 million displaced ones. **Tasks that remain human:** Strategic leadership; Physical trades; Healthcare delivery; Creative direction; Ethical governance **Tasks now at risk:** Data entry; Basic analysis; Template content; Routine coding; Document processing **Recommendation:** Industry statistics are useful for context, but your personal risk depends on your specific skill mix. Get a personalized score based on your actual resume. ## Reference table | Statistic | Value | Source | Year | | --- | --- | --- | --- | | Jobs exposed to AI automation globally | 300 million | Goldman Sachs | 2023 | | US jobs at high automation risk | 14% | OECD | 2024 | | Jobs displaced by 2025 (projected) | 85 million | World Economic Forum | 2020 | | New jobs created by AI (projected) | 97 million | World Economic Forum | 2020 | | Tech jobs lost in 2026 YTD | 100,000+ | Layoffs.fyi | 2026 | | AI CapEx by Big Four tech companies | $725 billion | 247 Wall St | 2026 | | Legal tasks exposed to generative AI | 44% | Goldman Sachs | 2023 | | Companies planning AI-driven workforce reductions | 41% | World Economic Forum | 2025 | | Workers needing reskilling by 2030 | 1 billion | McKinsey Global Institute | 2024 | ## Timeline - **2013** — Oxford study: 47% of US jobs at 'high risk' of computerization - **2017** — McKinsey: 400-800M workers displaced globally by 2030 - **2023** — Goldman Sachs: 300M jobs exposed to generative AI automation - **2024** — OECD revises: 14% at high risk, 32% facing significant change - **2025** — WEF: 41% of companies plan AI-driven workforce reductions by 2030 - **2026** — Tech layoffs hit 100K+ YTD; explicitly tied to AI restructuring ## The Numbers Behind the Headlines Every week brings a new headline about AI taking jobs. But the actual data tells a more nuanced - and more useful - story than the panic suggests. We've compiled the most important AI job replacement statistics from the world's leading research institutions. No cherry-picking, no exaggeration - just the numbers you need to make informed career decisions. ### The Big Picture: Global Impact **Goldman Sachs (March 2023)** published the most widely-cited estimate: roughly 300 million jobs worldwide are "exposed" to generative AI automation. This number sounds catastrophic - until you understand what "exposed" actually means. It refers to jobs where at least 50% of tasks *could potentially* be affected by AI. It does not mean those jobs will disappear. **The OECD (2024)** provides a more calibrated view: 14% of jobs across member countries face HIGH automation risk (meaning 70%+ of tasks are automatable), while another 32% face SIGNIFICANT changes (meaning 50-70% of tasks will be affected). That leaves 54% of jobs with relatively low direct automation exposure. **The World Economic Forum (2025)** estimates that while 85 million jobs may be displaced by automation by 2025, approximately 97 million new roles will emerge - a net positive of 12 million positions. However, the WEF also found that 41% of companies surveyed plan to reduce their workforce due to AI by 2030. ### The 2026 Reality Check The theoretical predictions are now meeting reality. Here's what's actually happening: - **100,000+ tech jobs eliminated** in 2026 as of mid-year - Layoff pace has accelerated to **~1,000 cuts per day** (vs. 674/day in 2025) - The "Big Four" tech companies have committed **$725 billion in AI CapEx** - For the first time, companies are **explicitly naming AI** as the reason for cuts - not "restructuring" or "efficiency" This represents a qualitative shift. In 2023-2024, layoffs were about post-COVID correction. In 2026, they're about permanent structural change. ### By Industry: Who's Most Exposed? The Goldman Sachs research broke down AI task exposure by sector: - **Administrative Support:** 46% of tasks exposed - **Legal Services:** 44% of tasks exposed - **Financial Operations:** 37% of tasks exposed - **Management Consulting:** 35% of tasks exposed - **IT & Software Development:** 36% of tasks exposed - **Healthcare Administration:** 28% of tasks exposed - **Healthcare Delivery:** 8% of tasks exposed - **Construction:** 6% of tasks exposed - **Maintenance & Repair:** 4% of tasks exposed The pattern is clear: **cognitive, desk-based, pattern-following work** is most exposed. Physical, emotional, and judgment-heavy work is most protected. ### What the Statistics Don't Tell You Here's the critical limitation of every statistic on this page: **they're all averages.** A "software engineer" has 36% task exposure according to Goldman Sachs. But a junior developer writing CRUD apps has 80%+ exposure, while a principal architect designing distributed systems has maybe 15%. Your risk depends on: - Your specific skills (not your job title) - Your seniority level - Your industry vertical - The tasks you actually perform daily That's why we built [Job Security Meter](https://jobsecuritymeter.com/) - to give you a *personalized* score based on your actual resume, not an industry average. ### The Reskilling Challenge Perhaps the most sobering statistic: **McKinsey estimates that 1 billion workers globally will need to be reskilled by 2030.** That's roughly one in eight working-age humans on Earth. The jobs that AI creates won't be the same as the jobs it displaces. They'll require different skills - AI orchestration, ethical governance, human-AI collaboration, and domain expertise that gives context to AI outputs. **Conclusion**: The statistics are clear: AI is transforming work at an unprecedented pace. But averages are misleading. [Upload your resume](https://jobsecuritymeter.com/) to get YOUR specific risk score - and a 6-month roadmap to get ahead of the curve. ## Questions ### How many jobs will AI replace by 2030? McKinsey Global Institute estimates that 400-800 million workers could be displaced by automation by 2030, but this is a range of technological capability, not a prediction. The World Economic Forum's more conservative 2025 report estimates 85 million jobs displaced but 97 million new ones created - a net positive of 12 million jobs. The real challenge isn't job loss but job transition. ### What percentage of jobs are at risk from AI? It depends on the study. The Oxford study (2013) estimated 47% at high risk. The OECD (2024) revised this to 14% at high risk and 32% facing significant changes. Goldman Sachs (2023) found that about 25% of current work tasks across all industries could be automated by generative AI. The most accurate answer is that virtually all knowledge work will be partially affected, but very few jobs will be fully replaced. ### Which industries will be most affected by AI? According to Goldman Sachs data, the industries with the highest task exposure are: Administrative support (46%), Legal services (44%), Financial operations (37%), and IT/Software (36%). However, exposure doesn't equal replacement - it means those tasks will evolve. Industries with the lowest exposure include Construction (6%), Maintenance (4%), and Healthcare delivery (8%). ### Is the 47% automation statistic accurate? The famous '47% of US jobs at risk' figure from the 2013 Oxford study by Frey & Osborne has been widely misquoted. The authors measured technological CAPABILITY, not actual job losses. The OECD later revised this dramatically downward to 14% at high risk. The original study remains valuable for understanding which tasks are automatable, but it should not be cited as a prediction of mass unemployment. ### How many tech jobs have been lost to AI in 2026? Over 100,000 tech jobs have been eliminated in 2026 as of mid-year, running at approximately 1,000 cuts per day. Major layoffs include Microsoft (125,000 cumulative), Meta (8,000 or 10% of workforce), Atlassian (1,600), and Cloudflare (1,100+). What's different about 2026 is that companies are explicitly citing AI efficiency as the reason - Cloudflare noted internal AI usage is up 600%. ### Will AI create more jobs than it destroys? Historical evidence says yes. The WEF projects 97 million new jobs vs. 85 million displaced - a net positive. But the catch is that the new jobs require different skills than the old ones. The transition period is painful for workers who can't reskill fast enough. McKinsey estimates 1 billion workers will need reskilling by 2030. The question isn't whether jobs will exist - it's whether you'll be qualified for them. --- **Source:** Job Security Meter — https://jobsecuritymeter.com/guides/ai-replacing-jobs-statistics-2026 **Human-readable version:** https://jobsecuritymeter.com/guides/ai-replacing-jobs-statistics-2026 **Last updated:** August 2026 **Attribution:** Free to quote and cite with attribution to Job Security Meter. **Full site specification for language models:** https://jobsecuritymeter.com/llms-full.txt