The Compression Exposure Index measures how exposed your role is to automation pressure, market compression, and structural pay shifts in the AI economy.
The Compression Exposure Index (CEI) is a 0 to 100 score that measures how fast a role's market value is being structurally compressed by AI automation, market re-pricing, and the redistribution of authority that the AI economy is producing. A higher CEI means faster compression; a lower CEI means a more durable structural position. It is the AI Career Index's central durability metric, expressed with a clear risk band so the reading is not abstract, and it is the number we use to triage every other reading in the assessment.
Compression is not the same as job loss. A role can be compressed long before it disappears. Compression means the market rewards your work less than it did, the scope of authority inside the role shrinks, the cost of replacing what you do with software falls below the cost of keeping you, and the wage premium your role used to command begins to leak into adjacent functions. That is the slope you want to measure before you are sliding down it. Most professionals notice compression in their pay packets a year or two after the structural conditions have already changed.
What compression actually looks like
Compression has a recognisable signature, and once you have seen it you start noticing it everywhere. Routine production work that used to take 90 minutes is now expected in 30. The number of analyst-hours allocated to a typical engagement gets cut in half between two budget cycles. Junior headcount on the team flatlines or contracts. The premium for being merely competent at the role evaporates and the premium for being the senior person who can supervise AI-assisted output expands. The work has not disappeared; the value capture has moved up the stack.
This is the macro pattern visible in every industry the AI Career Index tracks. The execution layer compresses fast. The judgement and accountability layer above it appreciates. The professionals who saw this clearly two years ago and moved are now in better positions. The professionals who waited for the compression to be unambiguous are now negotiating from a worse one. CEI is the early-warning instrument for that move.
What CEI actually measures
The CEI is built from a deterministic combination of structural inputs, weighted from your assessment answers. The five inputs are summarised below; their full definitions and weights are documented at methodology.
Routine cognitive load. The share of your weekly work that consists of producing specified outputs against known patterns. High routine load = high exposure. This is the single largest input to the score because routine cognitive work is the layer current AI tooling absorbs first.
Task boundedness. Whether your work is structured around discrete tasks with clear specifications or around outcomes with ambiguous paths. Task-bounded work compresses faster because each task can be templated, scored, and automated independently. Outcome-bounded work resists compression because the bridge from goal to execution itself does not template.
Value capture. Whether the economic value you produce is captured by you (through equity, performance fees, ownership) or by the employer (through fixed salary). Salary-only roles compress harder because the employer captures all the productivity gain when AI tooling lets the role do more, leaving the wage premium static while the work itself becomes cheaper.
Tool dependency. Whether your output runs through a single tool stack that itself is a substitution target. A senior performance marketer whose job is the orchestration of an ad platform that the platform itself is rapidly automating reads as high tool-dependency. A senior partner in a regulated profession whose work runs through her own judgement reads as low. Tool dependency adds risk because the tool's roadmap is now your career roadmap.
Industry exposure. The structural exposure of the industry your role sits in. Industries with heavy routine information work (call centres, BPO, basic content production, data entry) compress fast. Industries with regulated accountability layers (healthcare, regulated finance, licensed professional services), physical-presence requirements (skilled trades, in-person services, infrastructure), or judgement-density requirements (senior advisory, board-level work) compress slower. The wider AI Career Index dataset publishes hub-level readings for the 30 industry hubs so the industry coefficient is grounded rather than abstract.
Why CEI is deterministic
There is no AI guesswork inside the CEI. There is no probabilistic interpretation of your answers, no language-model reasoning over your inputs, no retrofitting against an outcome database. Every input maps to a number, every number maps to a weight, and every weight produces a reproducible score. The same answers always produce the same score. That is by design, not by limitation.
Structural scoring is only useful if it is reproducible. A career-durability metric that swings 15 points between two assessments of the same person is not measuring durability; it is measuring noise. The deterministic design is also what makes CEI auditable: a regulator, a board member, or a curious user can trace a score back to its inputs and see exactly which lever moved it. That property is rare in AI-era assessment tooling and it is the property the AI Career Index methodology is built around. The full audit trail is at methodology.
Reading your CEI score
A low CEI score does not mean your role is safe forever. It means the structural conditions of your role currently insulate you from compression at a measurable rate. A high CEI score does not mean you are about to be replaced. It means the structural conditions are changing what your work pays faster than the market average and your runway is shorter than it feels. The score is calibrated against the wider AI Career Index dataset, so a CEI of 45 sits in the moderate band whether you are a teacher, a financial analyst, or a software engineer.
The risk bands are intentionally simple. Low (0–35): structurally insulated, defending position is cheaper than rebuilding it. Moderate (36–65): bimodal exposure, where the next two years of moves matter more than the last five years of credentials. High (66–100): the substitutable layer is large, and the question is not whether to migrate but where to migrate to.
The point of the CEI is not to alarm you. The point is to give you a measurable starting position so that the moves you make over the next six to twenty-four months are calibrated to real structural pressure, not to news cycles or to the ambient anxiety around AI. The Compression Curve explainer walks through how compression unfolds over time once it begins, and the Acceleration Plan is the structured response built around your specific score.
Why CEI matters more than job-title risk lists
Generic at-risk lists rank entire job titles. A senior copywriter at a regulated bank and a senior copywriter at a venture-backed marketing agency share a job title and almost nothing else structurally. One is shielded by compliance friction and authority scope. The other is competing directly with generative output. A senior software engineer running a service at a top-five tech company and a senior software engineer rotating through tickets at a 50-person agency share a title and have radically different exposure. The CEI surfaces that distinction because it scores the structure of the work, not the label on the business card.
The AI Career Index publishes role-level CEI readings for the 1,014 roles in the framework, so you can see the role-baseline exposure even before you take the assessment. The personal CEI you receive from the assessment refines the role baseline against your specific structural inputs (your seniority, your value capture, your tool dependency, your industry context). The role-page reading and your personal reading often diverge by 10 to 20 points, and that gap is the most actionable single number in the report.
How CEI feeds the rest of the framework
CEI is the present-state reading. The Authority Band is the structural altitude reading. Together they form the two axes of the AI Career Index map. The 24-month projection (the Drift Path versus Strategic Path) is generated by running CEI forward under both no-intervention and intervention assumptions. The Acceleration Plan is built around the moves that change CEI fastest given your starting band.
Most professionals reading their CEI for the first time understate the score. The most common error is not over-reading the threat; it is under-reading the structural conditions. The two diagnostic questions that recalibrate it: What share of my week is producing artefacts versus making decisions? Could an experienced contractor on a quarterly retainer do most of the artefact work? If the answers are 'most artefacts' and 'plausibly yes', your CEI is higher than your gut suggests, and the value of running the actual assessment is highest.
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