How to Read AI Exposure Scores: What the Numbers Mean

PlainWorkforce assigns an AI exposure score to every occupation in the database. These scores are powerful tools for career planning — but only if you understand what they actually measure and, equally important, what they do not. This guide explains the methodology, the correct way to interpret the numbers, and common mistakes to avoid.

What the Score Measures

Each occupation in the O*NET database is defined by a detailed list of tasks — often 15 to 30 specific activities that workers in that role perform regularly. The AI exposure score evaluates each task against four criteria for AI automability:

• Routine data processing — Can AI handle this data entry, retrieval, or summarization task?
• Pattern recognition from structured data — Does this task involve identifying patterns, classifications, or anomalies in well-defined datasets?
• Rule-based decision making — Can this decision be expressed as a set of consistent, learnable rules?
• Predictable physical movements — Does this physical task occur in a structured, repetitive environment?

Tasks that match one or more criteria are scored as AI-susceptible. The occupation's overall score is the weighted proportion of its work content that falls into these categories. A score of 72% means 72% of the occupation's task weight involves activities that current or near-term AI systems could potentially perform.

What the Score Does Not Measure

This distinction is critical for avoiding misinterpretation:

• Not a job loss probability — A score of 85% does not mean 85% of workers will be laid off. Many occupations with high scores will experience task restructuring where AI handles routine components and humans focus on higher-value work.
• Not a timeline — The score measures technical susceptibility, not adoption speed. Regulatory barriers, cost economics, organizational inertia, and workforce dynamics all affect how quickly automation actually occurs.
• Not a quality judgment — A high score does not mean the job is "bad" or "unimportant." It measures a specific technical characteristic of the work content.
• Not company-specific — The same occupation can face very different AI adoption timelines depending on employer size, industry, and geography. A bookkeeper at a large tech company faces faster automation than one at a small rural business.

The Score Ranges

PlainWorkforce scores fall into interpretive bands that help contextualize the numbers:

• 0-20% (Very Low Exposure) — The occupation is dominated by tasks AI cannot perform well: complex physical work in unstructured environments, deep social interaction, or novel problem-solving. Examples: electricians, emergency medical technicians, psychiatric counselors.
• 20-40% (Low Exposure) — Most tasks are AI-resistant, with some automatable components. These roles will likely see modest AI augmentation in specific areas. Examples: registered nurses, school teachers, civil engineers.
• 40-60% (Moderate Exposure) — A meaningful share of work content is automatable, but substantial human-dependent tasks remain. These occupations will likely see significant restructuring. Examples: financial analysts, marketing managers, paralegals.
• 60-80% (High Exposure) — The majority of tasks are AI-susceptible. Workers should actively build skills in the non-automatable portions and monitor adoption in their specific role. Examples: bookkeepers, insurance underwriters, credit analysts.
• 80-100% (Very High Exposure) — Nearly all tasks can be performed by AI. These roles face the highest risk of substitution, though complete elimination depends on cost, regulation, and demand dynamics. Examples: data entry keyers, telemarketers, word processors.

Substitution vs. Augmentation

The most important nuance in reading AI exposure scores is distinguishing between substitution (AI replaces human workers) and augmentation (AI enhances human productivity). Both are consistent with a high exposure score, but their career implications are opposite.

Substitution is more likely when AI-susceptible tasks constitute the entire scope of a job — like data entry keying, where the full job is routine data processing. Augmentation is more likely when AI handles routine components while humans focus on judgment, relationships, or creative work — like a financial advisor using AI for portfolio analytics while spending more time on client relationships.

To assess which scenario applies to your role, look at the occupation detail pages and examine the specific task breakdown. If the non-automatable tasks are central to the occupation's value proposition, augmentation is likely. If the automatable tasks are the core of the job, substitution is the greater risk.

Combining Scores with BLS Projections

AI exposure scores are most useful when read alongside Bureau of Labor Statistics employment projections. The combination reveals four quadrants:

• Low exposure + Strong growth — The safest career zone. Healthcare clinical roles, skilled trades, and mental health counseling sit here. See the safest occupations ranking.
• High exposure + Strong growth — Augmentation zone. Software developers and financial advisors have high AI exposure but strong demand growth, suggesting AI augments rather than replaces. These roles require continuous skill development.
• Low exposure + Weak growth — Stable but stagnant. Some physical labor roles are AI-resistant but face limited demand growth due to other factors.
• High exposure + Declining growth — The highest risk zone. Data entry, word processing, and some clerical roles face both high automation potential and declining demand. Workers in these roles should actively plan transitions.

The fastest growing occupations ranking shows which roles combine the best of both dimensions.

Common Misinterpretations to Avoid

• "My job is 75% at risk" — No, 75% of your job's tasks are technically automatable. The actual risk to your employment depends on many factors beyond the score.
• "Low-scoring jobs are safe forever" — AI capabilities are expanding. An occupation with 15% exposure today could face 35% in five years as computer vision, robotics, and reasoning systems improve.
• "I should only consider low-scoring careers" — Some of the best-paying, fastest-growing occupations have moderate exposure scores. The key is whether the AI-resistant portions of the job are where value is created.
• "All workers in a high-scoring occupation face the same risk" — Within any occupation, individual workers differ in which tasks they emphasize. A high-scoring job title can be low-risk for an individual worker who specializes in the non-automatable components.

Using Scores for Career Planning

The most productive use of AI exposure scores is not to make binary stay-or-go decisions, but to guide continuous skill development. Look up your occupation, examine which tasks are flagged as automatable, and ask: Am I building depth in the tasks that AI cannot do? Am I learning to use AI tools to be more productive in my current role? Am I developing transferable skills that remain valuable even if my specific job title changes?

For detailed strategies, see our guide on how to future-proof your career.