The Research Behind the Score
CanIBeReplaced is not guesswork. Every scoring dimension is grounded in published workforce research from leading institutions. This page is updated as new studies emerge.
🔄 Last updated: April 2026Why Research-Backed Scoring Matters
Most AI job risk tools rely on theoretical task mapping — estimating what AI could do rather than what it is actually doing. CanIBeReplaced combines both: observed real-world AI usage data from Anthropic and Microsoft with adaptive capacity research from Brookings and NBER to give you a score that reflects both your current exposure and your ability to respond to it.
Studies Informing Our Scoring
8 published studies across labor economics, workforce research, and HR science
Labor Market Impacts of AI: A New Measure and Early Evidence
“The four sectors most exposed to AI displacement risk are Computer & Mathematical, Office & Administrative Support, Business & Financial, and Sales. At the other end, 30% of workers have zero AI exposure — jobs requiring physical presence that no LLM can replicate.”
Why it matters for your score: Our sector and environment scoring is directly calibrated against Anthropic's observed exposure data from millions of real Claude conversations.
Read the full study →Working with AI: Measuring the Applicability of Generative AI to Occupations
“Analysis of 200,000 anonymised Bing Copilot conversations found that knowledge work and communication-focused occupations face the highest AI applicability scores, while manual labour and physical roles ranked lowest.”
Why it matters for your score: Our task nature and digital exposure dimensions draw directly from Microsoft's real-world AI applicability scoring methodology.
Read the full study →Future of Jobs Report 2025
“92 million roles are projected to be displaced by 2030 while 170 million new roles emerge — a net gain of 78 million jobs. 41% of employers globally plan to reduce their workforce in areas where AI can automate tasks within the next five years.”
Why it matters for your score: Our tier labels and risk framing are calibrated against WEF's projected displacement timeline to ensure scores reflect realistic near-term risk.
Read the full study →State of AI 2025
“Today's technology could theoretically automate approximately 57% of current U.S. work hours — not 57% of jobs, but the hours worked across the population involving tasks a sufficiently deployed AI system could handle.”
Why it matters for your score: The distinction between task automation and job elimination underpins how we separate exposure risk from displacement certainty in our scoring.
Read the full study →Measuring US Workers Capacity to Adapt to AI-Driven Job Displacement
“Of 37.1 million highly AI-exposed workers, 26.5 million have above-median adaptive capacity. However 6.1 million workers face both high AI exposure and low adaptive capacity — concentrated in clerical and administrative roles, 86% of whom are women.”
Why it matters for your score: Our Future Readiness dimension — weighted at 30% of your total score — is directly informed by Brookings' adaptive capacity framework, which shows adaptability predicts outcomes better than exposure alone.
Read the full study →How Will AI Affect the Global Workforce?
“AI automation will displace roughly 6–7% of the U.S. workforce — approximately 11 million workers. Globally, around 300 million full-time jobs will be affected by generative AI.”
Why it matters for your score: Goldman's sector-level displacement modelling informs the risk weighting applied to financial, legal, and administrative role types in our scoring engine.
Read the full study →AI Jobs Barometer 2025
“Workers with demonstrable AI skills earn a 25% wage premium over peers without those skills. Median pay for AI-specific roles reached $156,998 in Q1 2025. Industries with higher AI adoption showed productivity growth rates four times higher.”
Why it matters for your score: The PwC wage premium data directly informs our recommendations engine — specifically why AI tool adoption appears as a high-priority action for most risk profiles.
Read the full study →Automation, Generative AI, and Job Displacement Risk
“Based on analysis of 20,262 U.S. workers, an estimated 63.3% of all wage and salary employment has at least one non-technical barrier preventing full automation — meaning human, legal, or contextual factors that AI cannot easily replace.”
Why it matters for your score: The SHRM non-technical barrier framework directly informs our Human Dependency and Emotional Intelligence dimensions — the factors that protect roles even in highly exposed sectors.
Read the full study →Key Statistics at a Glance
The numbers behind the risk landscape
of Computer & Math tasks theoretically automatable (Anthropic 2026)
of U.S. work hours involve automatable tasks (McKinsey 2025)
new roles emerging by 2030 (WEF 2025)
wage premium for workers with AI skills (PwC 2025)
of jobs have non-technical barriers to automation (SHRM 2025)
of workers have zero AI exposure (Anthropic 2026)
How This Research Shapes Your Score
Each dimension is weighted based on its predictive power in the research literature
Weights reflect the relative predictive strength of each dimension in published adaptive capacity and displacement research.
Stay Updated
The AI workforce research landscape moves fast. This page is updated as significant new studies are published. Take the assessment now to get your current score — and check back as the research evolves.
Take the Free Assessment →