How We Calculate
Your AI Risk Score
You're making career decisions based on our output, so you deserve to know exactly where the data comes from, how the scoring works, and where the limitations are. No black boxes.
Occupational Task Data — O*NET
When you upload your resume into the full AI Risk Assessment, AIRRBridge retrieves your correlating occupation's task list from O*NET OnLine, the primary occupational database maintained by the U.S. Department of Labor's Employment and Training Administration. O*NET documents the specific tasks, skills, and activities that define over 900 occupations across the U.S. economy.
What this means for you
Your risk score is not based on your job title alone. It's based on the actual documented tasks that define your role — the same task descriptions used by workforce economists, researchers, and policymakers to analyze labor market trends.
Update frequency: We have scheduled quarterly data updates even though O*NET itself updates its database annually. We just don't want to miss out on anything.
Real-Time AI Capability Assessment
Each task from your occupation's O*NET profile is evaluated against the current landscape of available AI tools and automation capabilities. This assessment is performed using large language model analysis — specifically evaluating whether existing, commercially available AI systems can perform, partially perform, or meaningfully augment each documented task.
Unlike static research reports that become outdated the moment they're published,AI capability is a constantly moving target. New tools enter the market monthly. Tasks that were low-risk in just last year may be highly automatable today. That's exactly why AIRRBridge is built as a subscription service — so your assessment reflects the AI landscape as it actually exists right now, not when you first signed up. Subscribers get access to updated assessments as our capability analysis is refreshed and our research grows, ensuring your career intelligence stays current as the market shifts beneath you.
What it evaluates
- →Can existing AI tools perform this task autonomously?
- →Can AI significantly reduce human time on this task?
- →Does this task require physical presence, emotional judgment, or novel reasoning that AI cannot replicate?
What it does not evaluate
- →Whether your specific employer will adopt AI
- →Labor market demand or job availability
- →Your individual performance or skill level
How the Score Is Calculated
Your AI risk score is calculated at the task level — not the job title level. Each of the 6–8 primary tasks documented for your occupation receives an individual automation susceptibility rating. Your overall score is a weighted composite of those individual task scores.
Example: How a score is built
What This Score Is — and Isn't
Being honest about limitations is part of being useful.
This is a capability assessment, not a job security guarantee
A high risk score means existing AI tools can perform a significant portion of your documented job tasks. It does not predict whether your employer will implement those tools, on what timeline, or whether your specific role will be eliminated. Labor markets are complex. The same score can mean different things in different industries, geographies, and organizations.
Scores reflect documented tasks, not your actual work
O*NET documents the typical tasks for an occupation. Your actual daily responsibilities may differ significantly from that profile. Two people with the same job title at different companies may have meaningfully different actual risk levels.
AI capabilities are changing rapidly
A task scored as low-risk today may become automatable within 12–24 months as AI capabilities advance. We update our assessments regularly, but no system can perfectly predict the pace of AI development. Use this score as a signal to act, not a permanent verdict.
Where We're Headed
Current state is the foundation. Here's what we're building toward.
Data Sources & Attribution
O*NET OnLine
Occupational task data sourced from the Occupational Information Network (O*NET), developed under sponsorship of the U.S. Department of Labor/Employment and Training Administration (USDOL/ETA). O*NET® is a trademark of USDOL/ETA.
onetonline.org →AI Capability Assessment
Task-level automation susceptibility assessed via large language model analysis of current commercially available AI tools and platforms, updated with each career page creation and quarterly thereafter.
Ready to see your score?
Now that you know how it works — take the free check. It takes 3 minutes and costs nothing.