AI exposure vs job replacement: what is the difference?
By the Mandlys editorial team · Updated
Exposure means AI can perform or affect some of the tasks in an occupation. Replacement means the job actually disappears. In between is augmentation, where AI makes people faster. Most major studies measure exposure, which is why high percentages do not mean that many people lose their jobs.
Three concepts that are often confused
- Exposure: how much of an occupation's tasks AI can technically perform or speed up.
- Automation: tasks actually taken over by machines, fully or partly.
- Augmentation: AI takes on part of the tasks so people can do more or different work.
The ILO's 2023 analysis found that the main effect of generative AI is likely to be augmentation rather than automation, and that clerical work is the most exposed group.
Why the numbers differ so much
Frey and Osborne (2017) assessed whole occupations and concluded that 47 percent of US employment was at high risk of automation. Arntz et al. (2016) redid the analysis for the OECD at task level and found 9 percent on average. The gap exists because most occupations combine automatable and non-automatable tasks.
Later studies such as Eloundou et al. (2024) and Felten et al. (2021) also measure exposure by task or ability, not actual job losses.
💡 Want to see how this applies to your own job? The free test takes about five minutes and needs no account. → Take the free AI test
What turns exposure into replacement
- Cost: whether AI is cheaper than the working time for the task.
- Quality and accountability: when mistakes are costly, human review is usually required.
- Rules and agreements: law, labour agreements and professional requirements.
- Share: if only a minority of tasks are exposed, the job changes rather than disappears.
How Mandlys uses these concepts
The Mandlys future index measures estimated exposure for your combination of tasks, role and development. It is not a probability of job loss. Read the full methodology.
Frequently asked questions
Does high AI exposure mean I will lose my job?
No. High exposure means many tasks can be affected. Whether the job disappears depends on cost, accountability, regulation and how large a share of tasks is affected.
Why did one study say 47 percent and another 9 percent?
Frey and Osborne (2017) assessed whole occupations. Arntz et al. (2016) assessed individual tasks within occupations and therefore found a much smaller share of high-risk jobs.
What is augmentation?
AI takes over or speeds up some tasks so people can do more or different work. The ILO considers it the most common effect of generative AI.
Sources
- Gmyrek, Berg & Bescond (2023). Generative AI and Jobs: A global analysis of potential effects on job quantity and quality. ILO Working Paper 96.
- Frey & Osborne (2017). The future of employment: How susceptible are jobs to computerisation? Technological Forecasting and Social Change, 114.
- Arntz, Gregory & Zierahn (2016). The Risk of Automation for Jobs in OECD Countries. OECD Social, Employment and Migration Working Papers No. 189.
- Eloundou, Manning, Mishkin & Rock (2024). GPTs are GPTs: Labor market impact potential of LLMs. Science, 384(6702).
- Felten, Raj & Seamans (2021). Occupational, industry, and geographic exposure to artificial intelligence. Strategic Management Journal, 42(12).