The team fears AI more than AI threatens it. At least for now.
I say this as someone who handed software development in his own company to AI agents and didn't let a single person go because of it.
A year ago, 32 in 100 companies surveyed by McKinsey expected AI to reduce their headcount within a year. A year passed. 14 in 100 reduced it. Two thirds of companies saw no significant change in the number of jobs. In every function asked about, the share of companies with actual reductions was smaller than the share that predicted them. In marketing, manufacturing, finance and knowledge management, by roughly half.
The fear, however, isn't going away. For the coming year, 39 in 100 companies expect reductions, more than a year ago. A similar group, 43 in 100, expects no change.
Eight in ten people in the same survey say AI has improved their own productivity, and half say it helps them make better decisions. Only 13 in 100 admit AI makes them anxious about their career. But there is a difference between levels: negative effects of AI at work are reported by 47 in 100 mid-level managers and individual contributors, against 31 in 100 executives. Those who decide about the deployment feel it the least.
What this means for a company that is only starting.
First: don't start the AI conversation with savings on headcount. The data says there most likely won't be any in the first year, and the fear you trigger will stay. Start with what people will feel themselves: less retyping, less searching through procedures, less answering the same questions.
Second: show the team where the saved time goes. At our company, the same team runs more projects at once, because agents do what is repetitive and people do what requires a decision. If you can't say what the team will do with the recovered hours, you are not ready for the deployment yet.
Third: start with a process where AI replaces nobody and only takes load off. A website chat that answers repetitive customer questions and hands the hard ones to a human is such a process. Nobody loses a job, and for the first time in a long while support has time for matters that require a head.
One caveat: this is data about the year that has passed. Companies are scaling agents faster and faster, and in two years the numbers may look different. That is why the time to get the team used to AI is now, while the risk to jobs is still small.
A question for you: has anyone in your company told the team what will happen to their work after the AI deployment?
Want to see how the agent handles your customers' questions? Enter your website address and e-mail in the demo form. In 20 minutes you get an agent trained on your website's offer, ready to be quizzed.
Sources
- McKinsey, The State of AI in 2026: On the road to ROI: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai