Methodology
Every occupation in our database is scored across 6 research-backed dimensions using data from 12 institutional sources. Here's exactly how it works.
Our AI risk score is a composite measure from 1 to 10 that synthesizes multiple research frameworks into a single actionable number. Each of the 997 occupations in the U.S. Department of Labor's O*NET database has been scored across 6 dimensions, each informed by peer-reviewed research and institutional data.
The dimensional scores were computed by mapping published research findings — automation probabilities, AI capability benchmarks, task-level exposure analyses, and labor market projections — onto standardized occupation codes. Three dimensions measure vulnerability (task automation, cognitive exposure, and the inverse of physical requirement), while three measure resilience (creativity, social intelligence, and regulatory barriers).
The composite score is a weighted combination where vulnerability dimensions push the score higher (more risk) and resilience dimensions push it lower (less risk). The weights reflect current AI capability trajectories: task automation carries the highest weight because it measures the most immediate and measurable displacement pressure.
Measures how many core tasks in this occupation can be performed by current or near-term AI systems. Based on O*NET task descriptions cross-referenced with demonstrated AI capabilities.
Evaluates how exposed the job's cognitive requirements are to AI substitution. Jobs requiring routine analytical thinking score higher (more exposed) than those requiring novel problem-solving.
Assesses the degree of physical, hands-on work involved. Higher physical requirements provide a natural barrier against AI displacement since robotics lags behind software AI.
Measures how much the role depends on original thinking, artistic expression, and novel idea generation. While AI can generate content, truly creative roles that require taste, vision, and cultural context remain resistant.
Evaluates the need for interpersonal skills, emotional intelligence, negotiation, and relationship management. Roles requiring deep human connection and trust remain AI-resistant.
Measures the extent to which licensing requirements, legal mandates, and regulatory frameworks protect the occupation from AI substitution. Some roles legally require human practitioners.
When you purchase a report, your job's pre-computed dimensional scores are combined with real-time data to generate a comprehensive, personalized analysis:
We use Claude (Anthropic) because it produces the most nuanced, well-reasoned analysis for complex career guidance. Each report is unique and generated in real-time — not a template.
997 occupation profiles with detailed task descriptions, skill requirements, and work context data. The foundation of our occupational taxonomy.
Median salary, employment count, and 10-year growth projections for every occupation. Provides the economic context for each role.
Pioneering 2013 study estimating automation probability for 702 occupations. Updated methodology applied to modern AI capabilities.
2023 research measuring which occupations and tasks are most exposed to large language models, providing task-level LLM exposure scores.
AI Occupational Exposure index mapping AI capability benchmarks to standardized occupation codes. Measures cognitive AI exposure per role.
Future of Jobs Report 2025 analyzing 55 economies, identifying which roles are growing, declining, or transforming due to AI adoption.
Analysis of 800+ occupations across 46 countries, estimating the share of work activities automatable with current technology.
Cross-country analysis of 27% average automation risk across G20 labor markets, with policy implications for workforce transition.
Assessment of 30,000+ tasks with 50,000 human evaluations, providing granular task-level automation vulnerability data.
2024 analysis finding 40% of global employment exposed to AI, with higher exposure (60%) in advanced economies.
Human-centered AI research tracking AI capability progress, economic impacts, and policy recommendations from Stanford University.
U.S.-focused analysis identifying 36 million workers in highly exposed occupations, with demographic and geographic breakdowns.
Search your job title and see your AI risk score instantly — for free.
Search Your Job →