Are employers ready to earn trust in AI?

Are we asking employees to train their AI replacements and calling it augmentation and empowerment?

In the short term, AI often does make people faster, more capable, and able to do things they couldn’t do before. But employees can do the same math executives can. If one person can do the work of two or three, at some point the company is going to ask whether it still needs two or three people. If an agent can learn a workflow, remember the exceptions, and keep improving, some of that augmentation starts to look like replacement.

I’ve been writing a lot about the value of a company’s memory: the customer history, the judgment calls, the workarounds, and all the things that never made it into a process document. Much of that knowledge lives in employees’ heads. Getting it into a form that AI can use requires those employees to explain what they do, why they do it that way, and when the usual rules don’t apply. The people you need to make the AI useful may reasonably believe you’re asking them to train their replacement.

I think employers need to be more honest about this, including admitting what they don’t know. AI may make your job dramatically better. It may change your role completely, eliminate parts of it, or eventually eliminate the job itself. It may also eliminate parts of management. Nobody can credibly promise that all of this productivity will arrive without changing how many people a company needs or what it needs them to do.

But uncertainty about the outcome doesn’t prevent a company from being clear about how it will treat people along the way. If someone helps automate half their job, what happens next? Do they get more interesting work? Does the company invest in helping them move into another role? Do they share in some of the value they helped create? Or have they just made the case for their own departure?

Employees will pay much more attention to what happens to that first person than to anything leadership says about its AI strategy. If helping the company adopt AI leads to a good outcome for that employee, others have a reason to participate. If it leads to a dismissal, the company has taught everyone else something too. You can’t ask people to share everything they know while giving them a reason to believe that keeping it to themselves is what protects their livelihood.

I don’t think the answer is to promise that AI will never replace anyone. Most companies probably can’t honestly make that promise. But they can be transparent about what they know, acknowledge what they don’t, and make real commitments about retraining, redeployment, and how productivity gains will be shared.

We talk a lot about whether employees are ready for AI. Maybe we should spend more time asking whether employers are ready to earn their trust by being honest about where this could lead, and admitting how much none of us knows.