A job title tells you almost nothing about its AI risk. How to read a posting for the parts AI can absorb versus the parts that hold, before you commit.
Most people evaluate a job by the title, the salary, and the company. Almost no one asks the question that matters most over a five-year horizon: how much of this role is work that AI is already learning to do? Two jobs with the same title and the same pay can sit at opposite ends of the AI-exposure scale, and you usually cannot tell which is which from the headline. This is a practical guide to how to check a job's AI risk before you apply, reading a job posting for its structural exposure, plus a free way to score it in seconds.
A job title tells you almost nothing about its AI risk
AI does not compress job titles. It compresses tasks. A role is exposed to the degree that its day-to-day work is routine cognitive production a model can reproduce, and it is durable to the degree that it rests on accountable judgment, ownership, and relationships a model cannot take over. Because the balance of those two layers varies enormously between roles that share a title, the title itself is close to useless as a risk signal. The same is true one level up at the industry level, which is why the reliable question is always about the actual work described in the posting, not the label on it. For the wider pattern of which roles are compressing and why, the guide to which jobs AI is replacing lays out the mechanics.
What actually decides a job's exposure
When you read a posting, you are really sorting its responsibilities into two buckets. The first is the substitutable layer: routine production that follows known patterns. The second is the durable layer: the accountable, judgment-heavy work that holds its value as the routine around it automates. The larger the first bucket relative to the second, the more exposed the role. These are the signals to watch for in the language of the posting itself:
- Substitutable signals: producing first-draft documents, formatting reports, data entry, scheduling, reconciling, processing, or summarising, and roles framed mostly around supporting and assisting.
- Durable signals: owning decisions, setting direction, managing people or budgets, advising clients, negotiating, and being accountable for an outcome rather than just producing an output.
- Authority language is the tell: words like lead, own, decide, and set strategy point to the durable layer; words like assist, support, coordinate, and process point to the substitutable one.
A good salary does not mean a safe job
It is tempting to read a high salary as a sign of safety, but pay reflects today's market, not tomorrow's structure. Plenty of well-paid roles are well paid precisely because they produce a lot of routine output quickly, and that output is exactly what AI is getting cheapest at. A posting can offer strong money and still be heavily exposed, just as a modestly paid role built on judgment and relationships can be far more durable. Read the structure and the salary as two separate questions, and do not let the number talk you out of the exposure check.
How to read a posting in a few seconds
You can sort a posting by hand, but the fastest way is to let the structure be scored for you. Paste the description into the AI job posting analyzer and it reads the language of the role and returns an AI-exposure score, an authority-layer band, and the five-dimension profile behind the AI Career Index, all on the same scale your own reading uses. It is free, needs no sign-up, and gives you a structural read of the job before you spend an afternoon on the application. Treat the number as a starting point, then sanity-check it against what you know about the team and the real scope of the work.
The questions to ask before you accept
A structural read is most useful when it turns into questions you put to the hiring manager. If a role looks heavily weighted toward the substitutable layer, you are not obliged to walk away, but you should find out what would move you toward the durable one. Ask:
- What decisions would I own here, and what am I accountable for, rather than just responsible for producing?
- How is the team already using AI, and would I be directing those tools or competing with them?
- Where does this role lead in two years: up the authority ladder, or sideways across more of the same production?
- Which part of this job would be hardest to hand to a capable AI tool today, and how much of my week would it be?
From the posting to your own position
Reading the job is half the picture; the other half is reading yourself against it. A role with real durable work only protects you if you actually occupy that layer rather than the production beneath it. To see where you sit and how a given move changes it, model your role in the AI Career Simulator, and browse how comparable roles are scored across the industries and roles library so the posting's number has context. The aim is not to chase the lowest-exposure job on the board, it is to choose work where your own judgment and ownership are the point.
A job posting is a description of where AI will and will not be able to take over the work, if you know how to read it. Sort the responsibilities into the routine layer and the accountable layer, weigh the balance, and ask what would shift you toward the part that holds its value. When you are ready to turn that into a plan for your own position, the AI Career Index assessment builds a sequenced six-month route from your own inputs.
Run your structural assessment
The AI Career Index is free to take. Get your structural snapshot in 10 minutes, including your durability score, archetype, and projection.
Take the Free Assessment