Everyone is debating whether AI will replace jobs. Meanwhile, your team already made up their minds, and they're acting on it whether you're talking to them or not.
That's the thread connecting five seemingly unrelated stories we've been watching this week. Canadian EV buyers are sitting on their wallets rather than settle for second-best. They've assessed the landscape, decided the leading option hasn't arrived yet, and chosen deliberate inaction over a purchase they'd regret. That's not apathy. That's a calculated bet. And your team is doing the exact same thing with their own effort and output, right now, inside your company.
Research from Nautilus found that workers who feel threatened by AI are more likely to disengage from their work. Not maliciously. Not dramatically. They just quietly pull back. Why build a skill you believe will be obsolete? Why go above and beyond for a company you think will automate you out? The rational response to perceived job extinction is to conserve energy. So that's what people do.
And here's the part founders get wrong: they think the danger is the AI tool. It isn't. The danger is the information vacuum around the AI tool.
Smashing Magazine's piece on botched AI adoption makes the same point the digital transformation era made, just with new vocabulary. Organizations fail at technology change not because they pick the wrong software, but because they treat the rollout as a technical problem when it's actually a cultural one. People don't resist tools. They resist ambiguity. They resist feeling like decisions are being made above them that will land on their heads. Give someone a clear role in a changing system and they'll adapt. Leave them guessing and they'll protect themselves.
We've seen this pattern across every serious technical engagement we've taken on over the years. The clients who botch internal tooling changes aren't the ones who pick the wrong stack. They're the ones who announce a new system and then go quiet. Two weeks later, half the team has found workarounds, the other half is waiting for it to be cancelled, and nobody is using the thing at full capacity. The vacuum fills with anxiety, and anxiety produces exactly the disengagement the Nautilus research describes.
Now layer in the labor law angle. Marginal Revolution's piece on regulatory mismatch makes an underappreciated point: having rules on the books isn't the same as having the organizational capacity to implement them well. Most companies that adopt AI have something like the same problem. They have policies, maybe even a formal AI governance document, but no actual operational muscle behind them. The rules exist. The culture to execute them doesn't. That gap is where things break.
And then there's the security dimension that almost nobody connects to the people problem. Forbes on proactive cybersecurity argues that the era of reacting to threats after they materialize is over. You have to build posture before the breach. The same logic applies to team culture around AI. Waiting until your best engineer is already coasting, or until your customer service lead has quietly started sending resumes, is reactive. By then you're doing damage control, not leadership.
So what does this actually mean for you?
If you're running a company between $1M and $5M, you almost certainly have someone on your team who is afraid. Maybe it's your ops person who knows you've been looking at automation tools. Maybe it's your designer who's seen what Midjourney can do. Maybe it's your developer who watches the AI coding tools get better every quarter. They're not bringing it up. They're not going to bring it up. But they're making decisions based on it every single day.
The EV buyer who sits on the sideline isn't irrational. They've read the room and decided the best play is to wait. Your threatened team members are doing the same thing. They're waiting to see if they still have a role. And while they wait, they're not building. They're not innovating. They're not doing the thing that actually makes your company worth anything.
The fix is not an all-hands meeting with a slide deck titled "Our AI Strategy." That will make things worse. People can smell performance from a mile away, and a town hall with bullet points is not the same as a real answer.
The fix is specificity. Tell your people what is actually changing, what is not changing, and what role they play in the new version. Not "we're exploring AI tools to improve efficiency." That sentence terrifies everyone in the room. Instead: "Here's the part of your job that's going to get easier. Here's the part we still need a human for. Here's where we need you to get better, and here's what we're going to do to help you get there."
Founders who are honest about this early don't just retain their team. They get something rarer: people who lean in instead of backing out. People who become the internal champions of tools they might otherwise have sabotaged. People who help you find the adoption path that actually fits your business rather than the one the vendor pitched.
The biggest AI risk in your company right now is not a data breach. It's not a bad vendor. It's the three good people on your team who have already mentally checked out because you haven't told them where they stand.
Talk to them first. Build second. Everything else follows from that.