What AI exposure actually means for your job

AI exposure is not a warning. It is a coordinate — a measured position on a 0 to 1 scale that describes how much of your current work overlaps with tasks AI systems are demonstrably capable of performing today.

That is the answer up front. The rest of this article explains where the number comes from, what it does and does not capture, and the practical moves that shift it.

Where the number comes from

Occupational exposure research maps job tasks to AI capability benchmarks — scoring tasks along dimensions like language processing, visual recognition, and reasoning, then aggregating across the task mix of an occupation. That published work established the approach.

cvgraph builds on it with its own capability-to-task mapping, kept current as AI capability moves, and applies the measurement to your personal skill and role graph — not to a generic occupation code. The difference matters: two people with the same job title can have meaningfully different exposure depending on which tasks they actually perform and which skills anchor their work.

The result is a number between 0 and 1. A score near 1 means most of your current task mix is in areas where AI capability is high. A score near 0 means most of your work is in areas where AI capability is currently limited.

What it does not capture

A single exposure score does not capture trajectory. It describes where you are, not where you are going.

It does not measure your value to an employer. Organisations need people who can direct, audit, and improve AI outputs — work that is complementary to AI capability, not competitive with it. High exposure on a narrow task cluster coexists with high value in a broader role.

It does not account for barriers to adoption — cost, regulation, trust, or the organisational inertia that slows even obvious automation. A high score in a regulated industry looks different from the same score in a fast-moving technology company.

Why it is a position you can move

Exposure scores respond to what you learn and what you build. Adding skills in areas of lower AI capability — human coordination, domain judgement, evidence synthesis, system design — shifts your aggregate score.

The loop in cvgraph makes this concrete: measure your current graph, identify the gap between your present position and a lower-exposure target, close the gap with deliberate skill work, then re-measure. The score is not a verdict handed down at one moment; it is a coordinate you can track across multiple points in time.

The re-measure loop

Re-measurement matters because AI capability is not static and neither is your graph. A skill you add this year changes your score. A capability boundary that AI crosses next year changes it again.

The useful question is not "what is my score today" but "which direction is my score moving, and am I moving it intentionally." The graph gives you the structure to answer that question with evidence rather than conjecture.

What to do with your number

If your score is high, the first step is clarity: which specific skills in your graph are driving it, and which are holding it down. That is what cvgraph's per-skill exposure breakdown surfaces — and when you are aiming at a specific role, JD-fit shows the concrete skill gaps between your graph and that job.

If your score is lower, the task is maintenance: staying current with how AI capability is expanding into the areas that currently protect your position.

In both cases, the score is a starting point for deliberate action, not a conclusion.