A manager has more job applications than she can possibly read. So she starts copying and pasting a large volume of resumes into a general-purpose AI tool and asking it to select five that best match the job description. She reads those five selected resumes properly. She's doing it in good faith, on top of her actual job, and her director thinks it sounds efficient.
Nobody decided she should do this. Nobody checked how the tool made its choices. And nobody read the other sixty.
An AI manager is a person who doesn't necessarily know how to build AI systems, but they know how to operate AI effectively within a business environment. The manager in our example has unknowingly become her company's AI manager.
Scenes like this are playing out in business everywhere. Around half of U.S. employees now use AI at work, according to recent federal and industry data. Most of that use happens the same way: one person solving one problem in one chat window. Most organizations already have AI managers. What they don't have is AI managers who know that's their job.
What is an AI manager?
An AI manager is the person who directs, checks, trains, and takes responsibility for the work AI does in their part of the business. Much like a traditional manager leads, develops, and oversees people, the AI manager simply applies the same thinking to AI.
If that sounds like management, that's because it is. A good manager briefs people clearly, explains what good results look like, checks the work, and owns the outcome. That's why the role is open to far more people than you might expect. You don't need to be a software developer to become an AI manager. You need to understand your role, your business, and the job at hand, and know how to direct AI to help you do it better.
Most organizations already have AI managers. What they don't have is AI managers who know that's their job.
If you already manage people, you have the foundations. If you don't, this is one of the clearest routes to more responsibility you'll find right now.

How do you become an AI manager?
Becoming an AI manager starts with a shift in thinking. Stop treating AI as just a piece of software and start treating it as something you manage. Then look at the work already on your desk and ask where AI could help without creating more work for you.
My book, The AI Manager, goes through this in detail, but there are three steps anyone can start with:
1. Give yourself permission to experiment: Try different AI tools in a safe, controlled space, and treat each one like a new team member. Instead of typing questions as you would into a search engine, practice briefing it as you would a colleague. Give it the outcome you need, the deadline, who it's for, and what it should flag if it runs into trouble. Then never accept the first answer. Make the AI prove itself. The more you experiment, the faster you will find your own style and see what it can really do.
2. Find opportunities to automate responsibly: Look for the repetitive tasks that eat into your week, like reports, summaries, first drafts, emails, or large amounts of data. Or, better yet, think about the problem you deal with every day that AI might solve. Be honest about the risk, too. If a task touches a decision about a person, such as who gets hired or who gets contacted, it needs far more care than a meeting summary.
3. Make it part of how you work: This is perhaps the most important step. Build what works into your everyday routine, so it saves you time rather than giving you another tool to manage. And keep checking it's working. Like any good manager, you'll learn when to trust the output and when to question it, and that judgment is what keeps the quality of your work high.
Why do businesses need AI managers?
For individuals, the case is simple. The working world is changing, and the people who learn to work well with AI will be the ones producing the best work and earning the next promotion.
For businesses, the stakes are higher. In 2025, the Chicago Sun-Times published a summer reading list that turned out to have been compiled with generative AI, and several of the recommended books didn't exist. A few months earlier, Coca-Cola had released an AI-generated remake of its classic "Holidays Are Coming" Christmas ad, and viewers were quick to criticize it as soulless and uncanny.
Neither was malicious. Both were simply AI projects that nobody was properly managing. More oversight would have prevented the brands from having to manage the unfavorable outcomes. That is why we need AI managers.
As AI becomes more capable, someone still has to decide what needs doing, explain what good looks like, provide the right context, check the output, and take responsibility for the final decision. That responsibility will become more formal. Before long, I believe every organization will need a named person accountable for how it uses AI, much as financial services firms already have named individuals accountable for how they're run.
So back to the question at the top of this newsletter. If you're using AI to get your work done, you're already managing it. The only choice is whether you do it deliberately.
Don't wait for someone else to step up. Take action to become the AI manager your organization needs, and help set the agenda for the next phase of automation.
Through the lens
Five to ten years' time is extremely exciting. Once we have navigated the issues currently being had with AI – like organizational change, dealing with data, and dealing with security – AI-driven workflows will take center stage.

Keep reading. It's free
- How do you become an AI manager?
- Why do businesses need AI managers?
- Through the lens
3 more minutes to read.
Read by 1,000+ founders & CEOs worldwide. No spam, ever. One click to unsubscribe.
