
AI Career Risk by Industry and Role
A clear read on AI job risk across 1592 roles, 31 industry categories, and 398 role-vs-role comparisons. Browse by industry hub or jump straight to your role.
A clear read on AI risk for every role we cover.
Every role page runs through the same five-part scoring method and produces the same set of results, so you can compare any two roles fairly. What changes is the role itself: the tasks it involves, how senior it is, how AI is changing its pay, and where you could move next if AI is taking over more of the work.
This is the full list, not a short curated pick. Browse by industry hub to see how a whole field stacks up on AI risk, or search and filter the list below to jump straight to one role. Each role page includes the risk read, typical pay from the US Bureau of Labor Statistics, task lists from O*NET, and the Acceleration Plan steps that keep the role strong over the next six months.
If your exact title is not listed, pick the closest match. The scoring looks at the work you actually do, not the job title, so two similar roles come out close. For side-by-side pairs, the role comparisons page is the better place to start.
Search 1592 roles, or browse by industry below.
31 Industry Hubs
Code generation, test automation, and routine debugging are the layers AI substitutes for fastest. The roles that thrive are the ones that own system design, security accountability, and the human-AI workflow boundary.
Campaign production, copy generation, performance reporting, and templated execution are directly substitutable by generative AI. Brand authority and customer insight are not.
Spreadsheet work, reconciliations, and report production are substitutable. Strategic advisory, risk underwriting, and accountability-bearing roles are not.
Screening, processing, and standard credit and compliance checks are substitutable. Deal judgment, complex underwriting, and accountability-bearing client and regulatory work are not.
AI substitutes efficiently for legal research and standard contract production. It cannot substitute for accountability, advocacy, or the personal trust at the heart of senior legal practice.
Workflow design, supplier negotiation, and process strategy retain value. Routine analysis, standard reporting, and templated coordination do not.
Trust-based relationships and strategic account ownership are not substitutable. Templated outbound, standard pipeline reporting, and routine follow-up are.
AI generates good-enough creative output at zero marginal cost. The roles that survive are the ones that own the creative direction the output serves.
Sourcing, screening, and scheduling are substitutable. Trust-bearing employee relations and strategic people decisions are not.
Computational design and BIM tooling compress drafting work. Site judgment, structural accountability, and licensure-bearing engineering decisions remain firmly human.
AI tooling absorbs listing production and standard valuations. The relationship and judgment work at the heart of property transactions does not.
AI absorbs routine coordination and reporting. Strategic supply chain decisions and supplier relationship work do not.
Strategic decisions, stakeholder accountability, and the personal authority required of executive leadership are structurally insulated.
AI lowers the operational cost of running a venture, which strengthens the economics of entrepreneurial roles. Ownership, strategic judgment, and execution accountability are structurally insulated.
AI augments diagnostics and absorbs administrative work. It cannot substitute for clinical accountability or the patient relationship.
Pedagogical authority and complex classroom work are durable. AI absorbs content production, standard assessment, and routine tutoring.
Physical accountability, on-site judgment, and complex process design are durable. Standard inspection and routine process work are partially substitutable.
AI absorbs routine merchandising, standard buying analysis, and templated ecommerce work. Senior retail leadership and complex brand-anchored work do not.
AI absorbs travel research and routine booking. Guest experience leadership and complex hospitality work do not.
Physical presence, on-site judgment, and accountability-bearing safety decisions are durable. AI absorbs routine monitoring and standard reporting.
Technical depth, on-site judgment, and regulatory accountability are durable. The category is in a structural growth phase.
Policy judgment, public accountability, and statutory roles are durable. Routine clerical, eligibility processing, and templated reporting are not.
Templated reporting, aggregation, and standard production are substitutable. Investigative depth, source relationships, and editorial judgment are not.
Standard bench work, routine data analysis, and templated review are substitutable. Original research design, peer judgment, and specialist expertise are not.
AI agents take the volume work in support and onboarding. The roles that thrive own enterprise renewals, executive relationships, and the design of the customer journey rather than the day-to-day of running it.
AI dismantles deck production and templated analysis, then exposes that judgment, sequencing, and stakeholder management is what clients actually pay for. The roles that thrive own engagement leadership and accountability, not throughput.
Visual production work compresses fastest as generative design takes over the surface. System design, accessibility, behavioural research, and design strategy gain value because they need judgment AI cannot carry.
Routine agronomic analysis, yield reporting, weather modelling, and standard scouting are directly substitutable by AI. Land judgment, animal husbandry, equipment operation, and farm-system architecture are not.
Routine intelligence analysis, supply chain administration, and standard staff work are being absorbed by AI-augmented defense platforms. Combat command, cyber operations leadership, and chain-of-custody decisions are not.
You cannot ship a haircut, a massage, or a one-on-one training session through a model. Personal-care work is bought because a person is physically present. Compression sits on the booking-and-admin layer, not on the service itself.
AI cannot crawl into an attic, hold a torch, or sign off on a permit. The skilled-trades workforce is protected by physical presence, licensure, and on-site judgment a model cannot deliver. The back-office (scheduling, dispatching, estimating) compresses; the tools do not.
All 1592 Roles
Search by role name, filter by risk band, or jump to an industry.
398 role comparisons
Careers people often mix up or weigh against each other. Filter and search them on the comparisons page.


AI Career Readiness Is the New Graduate Outcome
Employment rate and starting salary are lagging indicators. AI career readiness is the leading one, and universities that measure it will make better decisions about where to invest.

How to Choose a Career That Won't Be Automated
AI absorbs bounded production work first. The careers that hold up share four structural traits. Here is how to check any career path before you commit.
"The breakdown pointed out a gap between my skills and where the market is heading that I had kept overlooking. It is far more thorough than the usual career tools."
"I scored well on keeping up with AI, but my earning potential had a clear gap. The forecast made that problem easy to see and understand."
"The report showed me exactly where my weak spots were and what to work on next. Much more useful than generic career advice."
Measure Your Position with AI Career Index
Take the AI Career Snapshot to see your structural leverage, automation exposure, and growth trajectory with AI Career Index.
