AI Risk Scoring
How we evaluate career risk from artificial intelligence
The question we're answering
“What is the risk that AI will negatively impact this career — making it harder to keep your job or get hired after graduation?”
Time horizon: next 10 years — covers a high schooler through college and into early career.
The 1–5 Scale
AI cannot meaningfully perform the core work. The job fundamentally requires physical presence, human connection, or real-world manipulation that AI cannot replicate. Job numbers expected to hold steady or grow.
Examples: Electrician, Plumber, Firefighter, Carpenter, Dental Hygienist
AI assists with peripheral tasks but the core work remains human. Productivity gains are absorbed by growing demand or demographic need. Some workflows change, but total jobs remain stable.
Examples: Registered Nurse, Surgeon, Police Officer, Cybersecurity Analyst, Veterinarian
AI can perform meaningful portions of the core work. Employers will need fewer people to produce the same output. Entry-level hiring slows as AI handles junior-level tasks. Experienced workers adapt; new graduates face noticeably tighter competition.
Examples: Architect, Actuary, Civil Engineer, High School Teacher, Librarian
AI can perform most of the core work at near-professional quality. Significant headcount reductions expected. Entry-level roles are disappearing first. Graduates will compete for fewer openings against both humans and AI tooling.
Examples: UX/UI Designer, Data Engineer, Pharmacist, Paralegal, Journalist
AI already performs the core work at or above human level. The career is actively contracting. Most traditional roles will not exist in their current form within 10 years. Graduates should expect to pivot or specialize in a narrow niche.
Examples: Translator/Interpreter, Graphic Designer, Technical Writer, Paralegal
Scoring Principles
1. Task automation is the primary driver
If AI can do 70%+ of the daily work, that matters even if there's a current talent shortage — shortages close when fewer humans are needed.
2. Protective factors reduce the score but don't eliminate it
Licensing, physical presence, and regulation slow the timeline but don't stop it. A career where AI can do most tasks but licensing prevents replacement scores 3, not 1.
3. Entry-level impact matters most
If AI eliminates junior work first (and it usually does), that's especially relevant for students — even if senior practitioners are safe for now.
4. "AI as a tool" still reduces headcount
One person with AI doing the work of three means two fewer hires. Productivity gains are a risk to job numbers, not just a benefit.
5. Demand growth can offset automation
But only if it's structural (aging population needs more nurses) not cyclical. This is the main reason healthcare stays at low risk despite AI touching diagnostics and charting.
Important Notes
AI risk is not career quality. A score of 4 doesn't mean a career is bad — it means AI is changing the landscape significantly. Many high-risk careers (like software engineering) remain lucrative and fulfilling, but require adapting to AI-augmented workflows.
Scores reflect career-level trends, not individual outcomes. A talented, adaptable person can thrive in any field. These scores describe aggregate labor market shifts, not your personal ceiling.
We update regularly. AI capabilities change fast. Every career has a reasoning paragraph and cited sources so you can evaluate our logic and decide for yourself.
We err on the side of warning. It is better to warn you about a career that turns out fine than to reassure you about one that collapses. These scores are deliberately realistic, not optimistic.