While conversations about artificial intelligence are often dominated by fear of job losses, a new April 2026 report presents a more layered reality. This shows that while some parts of the workforce are highly exposed to automation, others remain firmly rooted in human capabilities.
The study by construction scheduling platform Planera focused on physical and manual labor, and the numbers revealed a clear difference. Emergency services have the lowest average automation risk at 11%, followed by social services at 12% and healthcare at 16%. These departments are much more difficult to replace because they rely heavily on human judgment, empathy, and real-time response.
Read: Why AI could be the biggest threat to home and remote working (
But more detailed data shows how risks are rising rapidly in other parts of the economy. Agriculture has the highest automation risk at 89%, followed by production at 82%, utilities at 81%, and retail at 80%. Logistics and mining both reach 77%, and transportation 68%. Even food services, which are often considered human-driven, face a 62% risk. The pattern is consistent. In other words, the more repetitive and process-driven the work, the easier it is to automate.
This trend becomes even clearer when we look at the individual roles most likely to be automated first. Metal and plastic pattern makers top the list with an automation risk of 99%, followed by underground mining equipment operators at 97% and milling and planing machine operators at 91%. Agricultural graders and sorters face 89% risk, and cashiers face 88% risk. While a wide range of production line jobs, from sewing machine operators to grinders and polishers, are in the mid-to-high 80% range, jobs such as post office sorters, meter readers and drivers continue to have risks above 75%.
Service roles are also unaffected. 71% of retail salespeople are at automation risk, compared to 69% of waiters and waitresses and 57% of restaurant cooks. The construction and maintenance landscape is even more complex, with some roles in the medium risk band, even though skilled trades have shown greater resilience.
Against this backdrop, the jobs that are most difficult to automate stand out even more. Emergency medical technicians are at the top of that list, with only a 7% risk. As first responders, they are required to assess medical conditions and provide life-saving care in unpredictable environments. The combination of urgency, physical skill, and decision-making keeps this role firmly human.
Firefighters follow at 9%, with increased strength across emergency services. The average risk for this sector is 11%, significantly lower than 37% for repair and maintenance and 38% for construction. Medical social workers are 12%, and empathy and communication remain central to their work.
Police officers and sheriff’s patrol officers also rank as the safest group at 13%. While some administrative tasks can be automated, core responsibilities still rely on human judgment and situational awareness.
Among industries, electricians stand out with a 14% automation risk and strong demand growth. The report highlights that this role sits at the intersection of low automation risk and increasing need, especially as infrastructure evolves.
Planera’s automation experts recognize this shift and say, “The combination of lower automation risk and increased demand is rare in today’s job market, but electricians, like many construction industries, have both. The electrician shortage is predicted to worsen through 2026, and the electrician shortage is expected to worsen through 2026, with aging infrastructure, More than 80,000 new jobs are expected across the country due to EV charging networks, and the energy transition. Ironically, the very technology that is driving concerns about automation, AI data centers, are the most future-proof jobs possible.
The study, which analyzed more than 55 manual workers, deliberately excluded office and technology roles to focus on the supposed physical backbone of the workforce. Researchers used employment data from the U.S. Bureau of Labor Statistics’ Occupational Employment and Wage Survey from May 2024 to narrow down the list of 63 roles across trade, production, logistics, health care, and services sectors. Automation risk estimates were derived from machine learning models trained on O*NET’s job attribute data to rank each occupation based on its likelihood of being automated.
What emerges is not just a warning, but a change in perspective. Jobs built on repetition are moving closer to large-scale automation and, in some cases, complete replacement. At the same time, the role that relies on human instinct, adaptability, and connection is not only constant, but becoming more important in an AI-driven economy.
This story is based on data shared by Planera.
