63% of roles requiring AI understanding are not in tech at all

If until now it seemed that the race for AI skills belonged mainly to programmers, software engineers, and data scientists, new data shows a completely different picture. Artificial intelligence is rapidly expanding beyond the boundaries of high-tech and entering the names of veteran and familiar roles – From human resource managers and marketers to physical therapists, instructors, salespersons, and drivers.

A new analysis by Indeed Hiring Lab, based on job postings in the US and five European countries, found that the number of role types in which employers integrated the term AI into the job title more than tripled in the US within four years. From 264 role types at the beginning of 2022, the number jumped to 822 in the first quarter of 2026. This represents 8.3% of the job titles included in the study – Roughly one in 12.

Yet the truly surprising figure lies outside of technology companies: 63% of the role types involving AI in the US already belonged to non-technological professions. A similar trend was recorded in Europe, where in more than half of the countries examined, most AI-related roles were outside the worlds of software and data.

To identify the trend, researchers examined job titles where at least five job postings during a given quarter included AI in the title. The choice to focus on job titles and not just the job description was meant to distinguish between an incidental mention of the technology and cases where the employer views it as a core part of the work or the required skills.

Among the postings located in the US were a test driver for an autonomous truck, a physical therapist who uses AI for documentation purposes, a real estate agent who works with an AI-based system to locate clients, and an electrical engineer for battery systems in artificial intelligence data centers.

An expansion into familiar professions was also recorded in Europe. In Germany, a position was posted for a human resource manager who is required to use AI to increase work efficiency; in France, they looked for salespersons for AI-based products and solutions; and in the Netherlands, jobs appeared for marketing and advertising personnel who use the technology for content creation and campaign management. The researchers identified three main areas in which new roles are being created or existing roles are being updated: Consulting and implementing AI in organizations, creating and testing content and data used to train systems, and training employees, clients, and students in using the new tools.

The meaning is not necessarily that hundreds of new professions were born overnight. In many cases, these are the same jobs that have existed for decades, but now a new layer of skills and tools has been added to them.

Pawel Adrjan, senior director of economic research at Indeed Hiring Lab and author of the analysis, explained to Business Insider that when AI appears in the job title, the data mainly reflects demand for workers who know how to integrate the technology into their work, rather than a signal that the employer plans to replace them. According to him, the demand is not necessarily for a computer science degree or deep technical understanding, but rather for a combination of professional familiarity in a specific field and the ability to work with AI tools.

A physical therapist still needs to know how to treat patients, a human resource manager is still required to understand people and organizational processes, and a salesperson still needs to know how to connect with clients. The difference is that employers expect them to use the new systems to document, analyze information, locate opportunities, or perform tasks much faster.

In this sense, AI may be undergoing a similar process to what computer skills went through in the past. In the first stage, they were considered an advantage for technological workers, and later became a basic requirement in almost every office and every profession. Now it seems that the ability to use artificial intelligence tools is beginning to take the same path.

For workers and job seekers, the message of the report is not necessarily that a full career change is required. A person who knows their field of work well is not required to become an AI engineer, but they may be required to explain how the new tools can help them perform their job better.

According to Adrjan, workers who can demonstrate familiarity with AI and explain how they use it in a professional framework are expected to be in a better position as this demand becomes explicit in job postings. Alongside this, he warned that the expansion of the demand could deepen gaps between workers who receive training and access to tools and those who fail to acquire the skills at the same pace.

Another analysis published by Indeed on the same day also points to a possible shift in the relationship between AI and employment. While between 2022 and 2026 the professions more exposed to artificial intelligence generally recorded a sharper decline in the number of jobs, over the past year it was precisely the most exposed professions that led the recovery in job postings. The researchers emphasize that it is still impossible to determine that AI is what created the recovery, but they raise the possibility that the technology is beginning to transition from a factor that reduces demand for workers to a tool that creates roles and changes existing jobs.

Either way, the direction is already clear: The question in the labor market is no longer just which professions will be replaced by AI, but which workers will know how to use it within the profession they already have.