Your Job Is Exposed to AI. Here Is How to Find Your Next Role.

If your job is exposed to AI, the realistic next move is almost never a leap from zero into something unrecognizable. It is a short, skill-adjacent hop into a role that reuses most of what you already know and carries lower AI exposure. Exposure is the start of a plan, not the end of a career. This article gives you a method you can run yourself this week: map your skill overlap with candidate roles, then rank those roles on four axes, namely skill overlap, target AI exposure, retraining time, and demand. None of it requires a leap of faith, and all of the data is public and free.

This is normal career maintenance, not a panic response. The World Economic Forum expects 39% of workers’ core skill sets to be transformed or become outdated between 2025 and 2030, down from 44% in its 2023 edition (World Economic Forum, Future of Jobs Report 2025). When more than a third of the labor market is re-skilling on a five-year cadence, planning a move is not a red flag. It is what staying current looks like.

JobRoute operationalizes exactly this routing per occupation across 1,016 O*NET occupations, reconciling the same public sources into one reproducible engine. You do not have to wait for that, though. The free tools described below let you start today.

What is an adjacent role, and why does a short hop beat reinvention?

An adjacent role is an occupation that shares the majority of your current skills, tasks, and knowledge. Most of your experience transfers, and the gap you have to close is small. That is the whole point: you keep the years you have already invested and redirect them, rather than discarding them.

This matches how the labor market actually behaves. The Brookings Institution, analyzing real worker transitions, found that job moves within the same skill-defined occupational cluster are 3.8 times likelier than moves across clusters (Escobari, Seyal, and Daboin Contreras, “Moving up”, 2021). People do not, as a rule, vault between unrelated fields. They step into nearby work that rewards what they already do well.

It helps to remember that exposure is measured, not guessed. If you want to see how exposure is scored before you plan around it, the formula and sources are public, so the number you are reacting to is something you can inspect.

How does skill-overlap mapping actually work?

Every occupation can be broken into components: the tasks it involves, the skills and knowledge it requires, and the abilities and work activities it draws on. Two roles are “adjacent” when those descriptor sets overlap heavily. That is the mechanism behind any credible adjacency map.

The U.S. public reference for this is ONET, the occupational data graph maintained by the Department of Labor. It describes 1,016 occupations using a content model of roughly 277 descriptors spanning tasks, skills, knowledge, abilities, and work activities (ONET Resource Center, U.S. Department of Labor, 2026). That structure is precisely what lets a tool say two roles share most of their skills.

These taxonomies are mature and standardized, which is what allows a statement like “role A and role B share 70% of their skills” to mean something concrete rather than being a guess.

Which free tools can you use yourself today?

You do not need a subscription to begin. Two free U.S. government tools give you a real adjacency map.

O*NET OnLine publishes a “Related Occupations” list on every occupation page, selected by shared characteristics, and flags in-demand roles as “Bright Outlook” (O*NET OnLine, U.S. Department of Labor, 2026). That is a built-in adjacency map for any job.

The BLS Occupational Outlook Handbook adds a free “Similar Occupations” tab to every job, listing roles that share duties, skills, education, or training, each with median pay and typical entry education (U.S. Bureau of Labor Statistics, OOH, 2026). That lets you compare candidate moves on pay and entry requirements side by side.

The honest limitation: you have to stitch these sources together by hand to answer “where should I go”. Each tool answers one part of the question. Closing that gap, in one reconciled view, is the work a routing engine does for you.

How do you weigh a candidate move across four axes?

The core framework is a four-axis scorecard. Run each candidate role through it.

1. Skill overlap: how much of what you know transfers. Favor roles that reuse the majority of your current skills. The Brookings data shows those within-cluster moves are 3.8 times more common, and therefore more achievable.
2. Target AI exposure: is the destination more durable. Aim for augmentation, not just difference. The clearest public signal of whether AI tends to assist or perform a task comes from usage data: on Claude.ai, augmentation slightly outweighed automation in November 2025, at 52% of conversations versus 45% (Anthropic Economic Index, January 2026 report). That figure reflects consumer Claude.ai usage; the same report notes that enterprise API traffic skews far more toward automation, so read it as one directional cue rather than a rule for all AI. Prefer destinations where AI assists a human who stays in the loop.
3. Retraining time: how big the gap is. A real move has a learning curve. Minimize it by choosing high-overlap destinations. For context, the WEF estimates 59 of every 100 workers will need training by 2030, with 29 upskilled in their current role, 19 reskilled and redeployed elsewhere, and 11 at risk of not getting the training they need, so a modest reskill is the norm, not a setback.
4. Demand: is the destination growing and paying. Route toward growth using BLS projections and Lightcast posting trends.

The labor market is churning, not shrinking. The WEF projects 170 million new jobs created and 92 million displaced by 2030, a net gain of about 78 million (World Economic Forum, Future of Jobs Report 2025). There are destinations to route toward.

What does a real routing example look like?

Here is a generic worked example you can copy, framed as a higher-exposure starting role moving to a lower-exposure adjacent destination that reuses the same core skills. It uses public O*NET and BLS categories for illustration only.

Start at your occupation’s page on O*NET OnLine and read its Related Occupations list. Shortlist two or three roles that share your core skills and carry a Bright Outlook flag. Then open each one in the BLS Occupational Outlook Handbook to compare median pay and typical entry education. Score the shortlist on the four axes.

The winning destination is Destination A: high overlap, a more durable augmentation-leaning task mix, a modest retraining gap, and a positive demand and pay signal. It is not Destination C, even though C “sounds” most unlike the old job. Novelty is not the goal; durability that reuses your experience is.

This example uses public categories for illustration. Per-occupation exposure scoring and adjacency are what JobRoute reconciles into one reproducible, version-locked engine, so you can see your adjacent roles in the AI Ready Score rather than assembling the picture by hand.

How do you tell whether a destination is actually more durable?

The trap is choosing a role that is merely different rather than genuinely more durable. A role can be unlike your current one and still be heavily automatable. Difference is not protection.

A practical durability test: look at whether the destination’s core tasks are ones AI tends to augment, meaning it assists a human who stays in the loop, versus automate, meaning it performs the task end to end. The Anthropic Economic Index found that on Claude.ai in November 2025, augmentation slightly outweighed automation, at 52% of conversations versus 45% (Anthropic Economic Index, January 2026 report). That is a consumer-platform signal rather than a measure of all AI, since the same report shows enterprise API usage skews toward automation, but it is a useful directional cue. Favor destinations whose core tasks lean toward the augmentation side.

A second cue: roles weighted toward judgment, interpersonal work, physical-in-context tasks, and accountability tend to be more augmentation-leaning. Both the WEF and the BLS point to human-plus-technical skill blends and care and healthcare roles growing, which is consistent with that pattern.

Pair a durable destination with durable personal skills. Knowing the durable skills to carry with you makes any adjacent move sturdier, because the skills travel with you regardless of which role you land in.

How do you check a target role is growing and pays enough?

Validate demand and pay before you commit, so the move is realistic and not just safer.

For demand, prefer ONET Bright Outlook roles, defined as occupations projected to grow 5% or more, or to have 75,000 or more job openings, over 2024 to 2034, or that are new and emerging (ONET OnLine, U.S. Department of Labor, 2026). For broader context, total U.S. employment is projected to grow 4.0% from 2023 to 2033, with healthcare among the fastest-growing groups: healthcare support occupations are projected to grow 15.2% and healthcare practitioners 8.6% (BLS Employment Projections 2023-33). Routing toward growing sectors stacks the odds.

For pay and openings, use the BLS Occupational Outlook Handbook’s Similar Occupations tab for median pay and entry education. For deeper wage detail, the BLS Occupational Employment and Wage Statistics program publishes employment and wage estimates for about 830 occupations (May 2024 release, U.S. Bureau of Labor Statistics). Cross-check both against Lightcast posting trends for what employers are hiring for right now.

What is the first step you should take this week?

Leave with momentum, not theory. Here is a sequence you can run in an afternoon.

1. Look up your current role on O*NET OnLine and read its Related Occupations list.
2. Shortlist two or three Bright Outlook destinations that reuse your core skills.
3. Open each in the BLS Occupational Outlook Handbook to compare median pay and typical entry education.
4. Score them on the four axes: skill overlap, target AI exposure, retraining time, and demand.

If you would rather have the routing done for you, the AI Ready Score reconciles O*NET tasks and skills, BLS wages and projections, Lightcast posting trends, and WEF and Anthropic exposure evidence per occupation into one reproducible, version-locked engine, so the four-axis comparison is done for you across 1,016 occupations.