Automation exposure indices measure the technical possibility that a technology could perform work — not whether it will, or whether anyone loses a job. Frey and Osborne put 47 per cent of US employment at high risk by scoring whole occupations. Arntz and colleagues got 9 per cent by scoring tasks instead. Eloundou and colleagues put 19 per cent of workers at more than half their tasks exposed to language models. None of them is a forecast.
Three studies, three answers, one question
Frey and Osborne scored 702 whole occupations for computerizability and put 47 per cent of US employment at high risk. Arntz, Gregory and Zierahn rebuilt the same question around tasks and got 9 per cent across OECD countries. Eloundou and colleagues measured exposure to large language models and found 19 per cent of workers with more than half their tasks affected.
Those are not three estimates of different things. They are three attempts at the same question, and they differ by a factor of five.
Why the unit of analysis decides everything
Scoring an occupation treats everybody holding that title as doing identical work. Scoring tasks recognizes that two people with the same job title spend their weeks differently, and that most jobs are bundles in which some parts are automatable and others are not.
When the analysis drops to tasks and allows for that variation, high-risk estimates fall from the high forties into single digits. That is not a refinement at the margin; it is most of the answer, and it is why the field settled on task-level decomposition as necessary rather than optional.
What each one actually measured
Frey and Osborne. Expert judgment on whether an entire occupation could be computerized. The results came out bimodal, with occupations clustering at the extremes, which is itself a symptom of a method too coarse for the question.
Arntz, Gregory and Zierahn. The same underlying question rebuilt around the task composition of jobs, across a set of OECD countries with comparable survey data.
Eloundou and colleagues. Exposure defined precisely and narrowly: whether a language model could cut the time a task takes by at least half while preserving quality. Note that this is a claim about speed, not about replacement.
They do not even agree on the ranking
Comparisons of these indices find that some correlate and some barely do. Measures built on machine-learning suitability and measures built on expert occupation scoring have shown little correlation despite claiming to describe the same phenomenon.
That is worth sitting with. If two indices rank the same occupations differently, at most one of them is describing reality, and there is no external test that tells you which.
The criticism worth carrying with you
The occupation-level scores track education level closely. That raises an uncomfortable question: are they measuring automatability, or are they measuring credential level and calling it automatability?
If it is partly the second, then using such a score to advise an individual is close to telling them their qualifications predict their risk, which is both less useful and less true than it sounds.
Why the numbers get quoted anyway
Because 47 per cent is a headline and 9 per cent is not. The high estimate came first, arrived with a striking number, and entered general circulation before the methodological response was published.
You will still see it quoted without the correction, frequently in serious places. That is not dishonesty so much as the ordinary lag between a finding and its critique, and the critique never travels as far.
How to read any exposure figure
Three questions. Occupation-level or task-level? Exposure to what specifically — robotics, software, language models? And over what horizon?
A number without those three attached cannot support a decision about your own work. With them attached it becomes usable, and usually more modest than the headline implied.
Common questions
What does an exposure score measure?
The technical possibility that a technology could perform work. Not whether employers will adopt it, and not job losses.
Why is 47 per cent so different from 9 per cent?
Unit of analysis. Occupation-level scoring treats every holder of a title identically; task-level scoring accounts for the fact that they do different work.
What did the 19 per cent figure measure?
The share of workers with more than half their tasks exposed to language models, where exposure means cutting task time by at least half without losing quality.
Do the indices agree with each other?
Not reliably. Some correlate and some barely do, which means at most one of any disagreeing pair describes reality.
How should I read an exposure number?
Ask whether it is occupation or task level, exposure to what exactly, and over what horizon. Without those three it cannot support a decision.
How much employment is at risk from automation?
Published estimates range from 47 per cent to 9 per cent depending on whether whole occupations or individual tasks are scored. There is no official figure.
Why is the range so wide?
Unit of analysis. Occupation-level scoring treats everyone with a job title identically; task-level scoring accounts for the fact that they do different work.
Do the indices agree with each other?
Not reliably. Some correlate and some barely do, which means at most one of any disagreeing pair describes reality — and no test tells you which.