The Bosses Using AI to Guess Who'll Quit Next — predictive…

Your manager might already know you're job hunting. They're just not telling you yet. A quiet shift is happening inside American companies, and it has nothing to do with the flashy AI chatbots everyone argues about online. It's predictive intelligence, and its target is your team. Software now scans calendars, emails, Slack messages, and even badge-swipe data to flag which employees are about to walk out the door, which teams are about to miss deadlines, and which managers are quietly burning out. If that sounds like science fiction, check your inbox. A growing number of HR departments at mid-size and large companies have already signed contracts for this kind of software. The pitch is simple: replacing exit interviews with exit predictions. Here's how it actually works. These tools pull together data you'd expect, like performance reviews and promotion history, plus data you might not, like how often you log in after hours, how quickly you reply to messages, and how many meetings you've declined lately. Then the system spits out a risk score. A sales rep in Phoenix might get flagged as high risk because her meeting load dropped 30 percent and she stopped volunteering for projects. A project manager in Ohio might be tagged as a retention concern because his team's internal chat activity went flat. Sounds efficient, right? That's the sales pitch. The reality is messier. For one thing, the accuracy is shaky. Most vendors admit their models catch some flight risks but also flag plenty of people who were never going anywhere. That means a loyal employee who simply had a slow month could get pulled into a private meeting with their boss and asked pointed questions about their future. Awkward doesn't cover it. There's also a legal minefield. If these tools use health data, family status, or other protected information, companies could run straight into discrimination claims. Some employment lawyers are already warning that a "high risk" score can become a self-fulfilling prophecy. Once a manager treats you like you're leaving, you start thinking about leaving. And then there's the trust problem. Employees generally don't know they're being scored. When they find out, usually through a leaked memo or a suspiciously timed conversation, the mood sours fast. One survey found that most workers say they'd be less loyal to an employer that used predictive tools on them without consent. So what should you actually do? First, don't panic. If you're doing good work and your manager treats you normally, you're probably fine. Second, watch for weird patterns. If your boss suddenly asks about your five-year plan or your commute, something might be up. Third, know your rights. If you're in a state with strong privacy laws, like California or Colorado, you may be able to ask what data your employer collects and how it's used. But the bigger point is this. Companies keep trying to solve retention with surveillance instead of, say, paying people better or letting them work from home two days a week. A risk score can tell you someone is unhappy. It can't tell you why. And it definitely can't fix a bad manager or a frozen salary. Our take: predictive intelligence isn't evil by itself. Used gently and transparently, it could help companies spot struggling teams before they collapse. But used in secret to rank and monitor workers, it's just another way to treat people like data points. If your employer starts predicting your exit, maybe it's time to make it a reality.