The fastest way to see what enterprise AI actually looks like in practice is to watch the layoffs.
GM just let go of about 600 IT workers, or more than 10% of the department, but the company isn't shrinking the team. It's hiring the same roles back, just for different skills.
This is a skills swap, not a cost cut - and it's exactly what most large companies are about to do.
The Old Job And The New Job
GM told TechCrunch it's still hiring in IT, but the job descriptions have changed. The company wants people who can build with AI from scratch, not just use AI tools to be more productive at their desks.
GM's posted skill list reads like an AI infrastructure team org chart:
- Engineers who build AI-native software
- Data engineering and analytics specialists
- Cloud engineers
- People running model development, prompt engineering, and agent workflows
In English: GM doesn't want IT workers who know the playbook from five years ago. It wants engineers who can train models, set up agents, and run pipelines that didn't exist before.
If you want a steady read on how AI is reshaping the workforce of every big company, Market Briefs covers it every weekday - plus a free investing masterclass when you sign up.
A Year Of Software Reshuffling
GM has been quietly tearing apart and rebuilding its software team for over a year.
In August 2024, the company cut about 1,000 software workers. In May 2025, GM hired Sterling Anderson - Aurora co-founder and longtime self-driving exec - as chief product officer.
By last November, three senior software execs were out: Baris Cetinok, Dave Richardson, and former chief AI officer Barak Turovsky, who held that job for just nine months.
The replacements lean hard into AI, with Behrad Toghi joining from Apple in October as AI lead, and Rashed Haq coming in as VP of autonomous vehicles after five years running AI and robotics at Cruise - the GM-acquired self-driving company GM later shut down.
The pattern is consistent: every new hire makes the team a little more AI-native, and every layoff round trims people whose skill set was built for an earlier version of GM software.
Worth Noting
GM is one of the biggest enterprise IT departments in America, so the shape of its hiring tells you what every large company is about to look like.
This isn't AI bolted onto a 2018 IT team. It's a different team.
Almost every legacy automaker has done some version of a software reset in recent years, but few have moved as quickly or as publicly as GM just did.
The investors who figure out which AI skills are actually scarce - and which companies are buying them in volume - are going to do a lot better than the ones still picking through a list of GPU stocks.
Join 350,000+ investors reading Market Briefs every weekday - and grab the 45-minute investing course that comes with it as a bonus.
