Transparent model
How we calculate the risk index
The risk index is an editorial estimate of technology exposure across tasks—not a forecast of layoffs. Future values are uncertain scenarios.
Formula
base = AI*0.30 + software*0.10 + robotics*0.15 + routine*0.15 + digitalization*0.10 + economic_incentive*0.20; index = base*(1-human_moat*0.005)
Reduces the final score to account for responsibility, empathy, trust, physical dexterity, unpredictable environments, negotiation, leadership and licensing.
What the index does not mean
- It is not the probability that an occupation disappears.
- It is not the probability that an individual loses their job.
- It is not an exact share of jobs that will be eliminated.
How the yearly outlook updates
The current year is detected automatically in the Yekaterinburg time zone. The site displays ten annual points: the current year plus nine more years. Values between anchors are interpolated, while years beyond the final anchor continue the latest observed rate and remain bounded to the 0–100 scale.
How explanations and skills are generated
Exposure reasons, protective factors, work tasks and skills combine a sector profile, role-specific keywords and risk-factor values. Every task percentage uses the relevant AI, software, robotics, routine and human-moat factors. On career-transition pages, transferable skills, gaps and action steps are generated separately for each occupation pair.
How the task quadrant is built
A task’s horizontal position shows the modeled technical ability of AI to perform or accelerate it. Its vertical position shows the need for a person, accounting for human moat, ambiguity and accountability. The quadrant explains the editorial model; it is not a separate laboratory benchmark.
Pay estimates
The Russian version uses modeled average monthly gross pay for Russia, anchored to Rosstat occupational and industry aggregates. The English version uses modeled average monthly gross pay for the United States, anchored to U.S. Bureau of Labor Statistics OEWS data. Occupation-level values are adjusted for sector, specialization, demand and the task-change scenario. The outlook through 2035 is nominal and does not guarantee specific income. The site does not estimate or display opening counts without verified data.
Evidence base
The ILO’s 2025 update mapped nearly 30,000 tasks and concluded that roughly one in four workers globally is in an occupation with some GenAI exposure. Transformation—not full human replacement—is the most likely effect.
The WEF estimated that structural change through 2030 may affect 22% of formal jobs, with about 170 million roles created and 92 million displaced. This covers technology, demographics, economics and other drivers—not AI alone.
Stanford AI Index 2026 describes uneven effects, with greater pressure in entry-level pipelines for some digital occupations. These observations cannot be generalized automatically across countries, age groups and roles.
Sources
Nearly 30,000 tasks; a global assessment of GenAI exposure across detailed occupational groups.
A survey of more than 1,000 employers representing over 14 million workers across 55 economies.
Evidence on the AI economy, productivity and uneven employment changes.
Occupational, task, skill and work-context structure.
International occupational classification.
Official aggregate earnings data for Russia, used to anchor modeled estimates rather than as an exact wage for every occupation.
Official U.S. occupation-level employment and wage estimates used to anchor the English-language market model.
Official long-term U.S. employment projections used as an external signal for demand direction.