AI’s Real Impact On Jobs Is Still Unfolding

Will AI take your job? It’s a question many workers and students are asking as AI tools become more common in offices and industries around the world. A recent labour market study by the AI company Anthropic offers some insight into how AI might affect different professions.

The study examined nearly 800 occupations and analysed the tasks involved in each job. Researchers compared those tasks with what modern AI systems, including tools like Claude AI, are capable of doing. A central point in the study is the gap between what AI could do in theory and what it is actually doing at work today. Researchers looked at the tasks that make up different professions and compared them with the kinds of work AI tools are capable of handling. They found that the technology may be able to perform a large portion of certain jobs, but in most workplaces that level of automation hasn’t happened yet.

Take computer- and mathematics-related jobs as an example. According to the analysis, AI systems could potentially handle as much as 94 percent of the tasks involved in these roles. In reality, only about a third of that work is currently being supported or carried out by AI tools. In office administration, the change is already becoming visible, with nearly 40 percent of routine tasks now supported by digital tools. The study also highlights certain jobs where routine work makes it easier to introduce new digital tools.

Professions such as computer programmers, customer service staff, data entry operators, and financial analysts fall into this category. Much of their work involves going through large amounts of information, preparing reports, responding to queries, and checking data, which means some parts of these roles can be handled more efficiently with the help of digital systems.

The impact isn’t uniform across industries. Sectors like construction, agriculture, protective services, and personal care rely on hands-on work and direct interaction, making them much harder to automate. For now, the findings suggest that AI is becoming more common in some workplaces, but its use is still far from reaching its full potential. The research also shows that its impact may first appear in hiring patterns rather than layoffs.

Although the research looks mainly at the United States, the findings are relevant to other countries too. In India, concerns have already begun to emerge in the IT services sector. At the same time, economists point out that predicting how jobs will change has never been straightforward. In earlier waves of technological change, entirely new kinds of work appeared—roles that very few people had imagined at the beginning. For now, the most honest answer to the question “Will AI take your job?” may simply be this: No one knows yet. But the workforce of the future will likely belong to those who keep learning, adapting, and evolving alongside the technology.