Artificial intelligence may be reshaping how people work, but the large-scale job losses many feared have yet to materialise. A new report from Anthropic attempts to map where the biggest risks could emerge in the future by identifying professions most vulnerable to AI-driven automation.
The research introduces a new framework to measure what the company calls “AI displacement risk”, essentially estimating which occupations could see tasks increasingly handled by artificial intelligence systems.
The AI firms, which have recently been seen in the ruffle with the Pentagon, explain the approach in cautious terms, emphasising that it is an early attempt to track the labour market impact of AI.
“In this paper, we present a new framework for understanding AI’s labor market impacts, and test it against early data, finding limited evidence that AI has affected employment to date. Our goal is to establish an approach for measuring how AI is affecting employment, and to revisit these analyses periodically.”
The report arrives at a time when AI tools are rapidly becoming embedded in workplaces, from customer service chatbots to advanced coding assistants.
10 jobs at highest risk
According to the analysis, the occupations facing the highest exposure to AI are largely concentrated in information-heavy or repetitive knowledge work. These are roles where tasks involve processing text, generating content, or analysing structured information, areas where modern AI systems already perform strongly.
The paper states, “In line with other data showing that Claude is extensively used for coding, Computer Programmers are at the top, with 75 per cent coverage, followed by Customer Service Representatives, whose main tasks we increasingly see in first-party API traffic. Finally, Data Entry Keyers, whose primary task of reading source documents and entering data is significantly automated, are 67 per cent covered.
Among the jobs identified in the report are:
- Computer programmers
- Customer service representatives
- Data entry clerks
- Legal assistants and paralegals
- Proofreaders and copy editors
- Technical writers
- Market research analysts
- Administrative assistants
- Translators
- Content writers
Many of these roles involve structured tasks such as writing standardised documents, responding to customer queries, analysing survey data, or editing text. AI systems are increasingly capable of performing these functions quickly and at scale.
However, the report does not suggest that these jobs will disappear overnight. Instead, it indicates that specific tasks within these professions could be automated, potentially changing how the work is performed rather than eliminating the roles. For example, AI may draft documents or summarise information, while humans still review outputs, make final decisions, and manage complex cases that require judgement or empathy.
How does Anthropic measure work?
The company’s analysis examines three main elements: the types of tasks that make up a particular job, how many of those tasks could, in theory, be handled by large language models, and how many are already performed by AI tools in real-world settings.
Roles are considered more vulnerable when a large share of their core responsibilities can be automated and when AI systems are already taking over those functions in practice.
The researchers behind the study also acknowledge that the model has limitations. They note that the methodology may not capture every possible way AI could transform jobs and industries. However, the team argues that establishing a framework now, before major disruptions occur, will make it easier to detect real economic changes in the future rather than analysing them after the fact.
AI vs human jobs
The debate over whether AI will replace human workers has been ongoing since ChatGPT launched in 2022, triggering a wave of interest in generative AI tools.
Dario Amodei, the chief executive of Anthropic, has repeatedly said the reality is more nuanced than a simple case of job replacement. While AI is already outperforming humans in some narrow technical tasks, it has not yet led to widespread job losses.
Citing an example from the medical field, he pointed to predictions made by AI pioneer Geoffrey Hinton.
“Geoff Hinton predicted that AI will replace radiologists… and indeed, AI has gotten better than radiologists at doing scans.”
Yet the expected collapse of the profession has not happened.
“There aren’t fewer radiologists,” Amodei added. “The most highly technical part of the job has gone away, but somehow there’s still demand for the underlying human skill.”
Still, the balance between humans and machines could change over time as AI capabilities continue to evolve.
“Perhaps, over time, AI will advance in areas where it hasn’t yet advanced, and maybe that’ll happen fast,” he cautioned. His approach remains empirical: “We should take it one step at a time and see what AI does today.”
For now, the report suggests that AI’s biggest impact may not be replacing workers outright but quietly transforming how many jobs are done.
