TechCerted Index · 2026
The AI-Proof
Career Index
We scored 18 tech careers from 0 to 100 on how resistant they are to being displaced by AI over the next five years. One transparent score, four measured factors, real salary data.
Safest from AI
Cybersecurity Analyst
AI-Proof Score 94/100
Most exposed
Prompt Engineer
AI-Proof Score 42/100
The evidence base
What the research actually says
We didn't invent the direction of these scores. Every major study of AI and work points the same way: routine cognitive, analytical, and administrative tasks are the most exposed, while judgment-heavy, interpersonal, and non-routine work is the most resilient. Our four factors are built to mirror that consensus.
“Even where tasks are exposed to AI, only about a quarter are cost-effective to automate within a decade.”
Which U.S. Workers Are More Exposed to AI on Their Jobs? · 2023
Exposure tracks routine cognitive tasks — “getting information” and “analyzing data.” Jobs that care for people or do physical work are the least exposed.
Read the sourceGenerative AI Could Raise Global GDP by 7% · 2023
Up to 300 million full-time jobs are exposed to automation. Office/administrative support (~46% of tasks) and legal (~44%) carry the highest task-automation share.
Read the sourceThe Simple Macroeconomics of AI · 2024
Even where tasks are exposed, only ~23% are cost-effective to automate within a decade — so role-level displacement is gradual and partial, not a switch.
Read the sourceFuture of Jobs Report 2025 · 2025
170M jobs created, 92M displaced by 2030 — a net positive, but with churn equal to 22% of all jobs. The question is which roles, not whether jobs.
Read the sourceGenerative AI and the Future of Work in America · 2023
Up to 30% of hours worked could be automated by 2030. Office support, customer service, food service and production roles see the largest demand decline.
Read the sourceGen-AI: AI and the Future of Work · 2024
In advanced economies ~60% of jobs are exposed to AI — split roughly half augmented, half at displacement risk. Higher-skill white-collar work is most exposed.
Read the sourceHow we score it
The methodology
Each career gets four factor ratings (0-100) from our editorial model. The two heaviest factors — judgment and automation resistance — carry 30% each because that is exactly where the research above (Pew, Goldman, Acemoglu) locates the dividing line between exposed and resilient work. The AI-Proof Score is a fixed weighted blend of all four, so the headline number is always reproducible from the published inputs. No black box.
Human judgment & ambiguity
How much the work depends on non-codifiable judgment, novel problems, and decisions with no clean right answer. AI is weakest here.
Automation resistance
The inverse of how much of the day-to-day is exactly what today's AI already does well (generating code, drafts, dashboards, mock-ups). Lower if the core tasks are already automatable.
Accountability & risk ownership
Whether the role carries human accountability — legal, financial, or safety liability — that organizations will not hand to a model.
Stakeholder & human interface
How much the job is persuading, coordinating, and leading people across an organization — work that stays human even as tools change.
Methodology v1.0 · 2026. Salary and growth figures from the TechCerted salary dataset. This is an editorial model for guidance, not a prediction of any individual's job security. Citations welcome — credit “TechCerted AI-Proof Career Index” with a link to this page.
Move toward an AI-proof career
The safest roles all reward a credential that proves you can do the judgment-heavy work. See the fastest route in.