How do I know if my job is at risk from AI?

According to Dr. Seungbin Yim's research framework, evaluate your job on 5 criteria: (1) Routine cognitive content — what percentage of your work follows predictable patterns AI can learn? (2) Judgment complexity — does your role require contextual decisions that need human experience? (3) Human interaction depth — does your work require emotional intelligence, trust-building, or persuasion? (4) Physical presence requirement — must you be physically present to perform the work? (5) Adaptation speed — how quickly does your field's knowledge base change? Jobs scoring high on routine content and low on the other four face the greatest risk.

What careers are AI-proof?

No career is fully AI-proof, but certain roles have strong structural resistance. Jobs that combine complex judgment, deep human interaction, and rapid adaptation requirements are most resilient. Examples include: crisis management professionals, investigative journalists, specialized therapists, skilled trade supervisors, and roles requiring novel creative problem-solving. The key insight from research is that AI-resistance is not about the job title but about the specific task composition within each role.

What is the AI Class Society?

The AI Class Society is a concept from Dr. Seungbin Yim's book of the same name, describing how AI is creating a new socioeconomic stratification. The labor market is splitting into three tiers: AI creators (those who build and control AI systems, capturing disproportionate value), AI operators (those who use AI tools effectively but depend on them), and AI-displaced (those whose skills have been automated). Data shows AI hiring is up 92% with a 56% wage premium, while administrative roles have declined 21% — indicating this stratification is already measurable.

Assessment Framework

AI Career Risk Assessment: 5 Criteria That Determine Your Job's Vulnerability

A research-based framework for evaluating which careers survive AI disruption — and which ones are already eroding.

By Dr. Seungbin Yim · Updated May 2026

The 5-Criteria Framework

Not all jobs face equal risk from AI. The difference between an AI-resilient career and a vulnerable one can be measured across five dimensions. This framework, developed from analysis across multiple research institutions, provides a structured way to assess any role.

Routine Cognitive Content

Question: What percentage of your daily work follows predictable, pattern-based processes?

Roles where 70%+ of tasks follow repeatable patterns face the highest automation risk. Data entry (95% routine), basic accounting (80%), standard legal review (75%) are in the danger zone. Research shows AI excels at pattern recognition but struggles with novel situations.

Judgment Complexity

Question: Does your role require decisions that depend on context, experience, and incomplete information?

Roles requiring complex, contextual judgment — such as crisis management, strategic planning, or diagnostic medicine — have high AI resistance. The key differentiator: can the judgment criteria be fully codified, or does it require human experience and intuition?

Human Interaction Depth

Question: Does your work require trust-building, emotional intelligence, or persuasion?

Roles involving deep human interaction — therapy, negotiation, leadership, teaching — maintain resistance because the interaction itself is the value, not just the information exchange. AI can simulate empathy but cannot authentically build trust.

Physical Presence Requirement

Question: Must you be physically present to perform the core value of your work?

Skilled trades, emergency response, surgical procedures, and hands-on creative work maintain structural resistance. Currently less than 5% of physical-presence tasks are automatable by AI.

Adaptation Speed

Question: How quickly does the knowledge base in your field change?

Paradoxically, fields with rapid knowledge turnover can be either more or less AI-resistant. If AI can be trained on new data faster than humans can learn, the field becomes vulnerable. If the field requires creative integration of new knowledge with existing expertise, human advantage persists.

Risk Matrix by Occupation Category

Occupation	Routine %	Judgment	Human	Physical	Risk Level
Data Entry / Processing	95%	Low	Low	No	Critical
Basic Accounting	80%	Low	Low	No	High
Customer Service (Script)	85%	Low	Med	No	High
Junior Software Dev	60%	Med	Low	No	Medium-High
Marketing Analyst	55%	Med	Med	No	Medium
Project Manager	40%	High	High	No	Medium-Low
Therapist / Counselor	15%	High	High	Partial	Low
Skilled Electrician	20%	Med	Med	Yes	Low
Emergency Physician	10%	High	High	Yes	Very Low

The Three-Tier AI Class Structure

AI is not just changing jobs — it's creating a new class system. Research shows the labor market is stratifying into three distinct tiers, with declining mobility between them.

Tier	Role	Market Signal
AI Creators	Build, architect, and control AI systems	+92% hiring, +56% wage premium
AI Operators	Use AI tools effectively as productivity multipliers	Stable employment, wage stagnation
AI-Displaced	Skills automated or value hollowed by AI	-21% roles, downward pressure

The critical insight: most professionals currently sit in Tier 2 (Operators) and assume stability. But the boundary between Operator and Displaced is shifting upward as AI capabilities expand — what required an Operator last year may only need a prompt this year.

Assess Your Own Risk

Take the free self-assessment or explore the full research in the book series.

Free Career Risk Assessment → Read Quiet Erosion