Calling HR a “cost center” has always been a lazy way for leaders to dodge responsibility. It’s a neat label that lets you underinvest, then complain that HR is slow, rules-obsessed, and out of touch. So when I see people arguing that AI is the chance to turn HR into a driver of business impact, I’m torn: the idea is right, but the way it gets sold can turn into a new kind of corporate fantasy.
Here’s the basic claim, from what’s been shared publicly. HR has been stuck in a traditional service model. Lots of tickets, forms, approvals, handoffs. “Service delivery” becomes the job, not building a workforce that actually helps the business win. That model creates inefficiency and disconnects HR from real outcomes. The proposed fix is an “Intelligent Service Delivery” model with three pillars: Intentional Experience, Connected Intelligence, and Embedded Impact. And AI is positioned as the thing that makes this shift possible—better service, better decisions, more measurable impact.
I buy the criticism of the old model. If HR is mainly a help desk plus a policy gate, it will always feel like friction. Employees experience HR as “the people who tell me no” or “the people who need another form.” Managers experience HR as “the people who slow hiring down.” And HR teams, honestly, get trained into defensive habits: reduce risk, follow process, close the ticket. It’s not evil. It’s what you get when the scorecard is speed and compliance, not whether you built a strong team.
But here’s my problem: swapping “service delivery” for “impact” can become a rebrand that hides the same old issues. The biggest limit in HR isn’t that the tools are too dumb. It’s that the business often wants HR to be a shield. Leaders say they want “better workforce decisions,” but what they actually want is cover for decisions they’ve already made. AI won’t fix that. It might even make it worse, because a machine-generated answer can look like truth when it’s really just a cleaner-looking opinion.
Take “Intentional Experience.” On paper, great. Employees shouldn’t feel like they’re navigating a maze to get a simple answer about pay, leave, or performance. But if “experience” becomes a design exercise while the underlying policies stay hostile, it’s lipstick. Imagine an employee who’s trying to take time off for a family emergency. A slick AI chat system can respond fast, sure. But if the policy is rigid and the manager is penalized for approving time, the “experience” is still bad. Now it’s just bad faster.
“Connected Intelligence” is where the stakes jump. This sounds like joining data across the employee lifecycle so HR and leaders can make smarter calls. Again, good idea. Decisions about hiring, promotions, pay, and layoffs should not be vibes-based. But connected data is also a temptation. If you can measure everything, you will start treating people like variables to optimize. And once leadership gets a dashboard, they’ll start asking for simple answers to messy questions: “Who is low performance?” “Who is a flight risk?” “Which teams are overpaid?” Those aren’t neutral questions. They’re loaded. The wrong answer doesn’t just hurt a metric; it can wreck someone’s career.
Imagine you’re a manager and you’re told the system flags one of your team as likely to quit. What do you do? Do you invest in them, or do you quietly stop giving them big projects because you don’t want to bet on someone “at risk”? A tool meant to help retention can become a self-fulfilling push out the door. That’s not a tech problem. That’s incentives and human nature.
“Embedded Impact” is the part I agree with most—and the part that could go most wrong. The pitch is that HR shouldn’t just process work; it should shape execution. HR should be in the flow of decisions. That sounds like progress. But “embedded” can also mean “always watching” or “always enforcing.” If HR gets closer to day-to-day business moves, it can either become a partner that makes teams healthier, or it can become a control layer that turns every decision into a compliance ritual.
The shift from “efficiency” to “quality of workforce decisions” is the right direction, but it’s also dangerous because it invites overconfidence. Quality is hard to define. If the business says quality means “move faster and cut cost,” HR will use AI to do exactly that. If quality means “fairness and long-term capability,” HR will push different choices that might be slower and more expensive in the short term. AI doesn’t pick the values. Leaders do.
And that’s where I land: this “HR as impact driver” moment is promising only if it forces leaders to own what they’re optimizing for. If the goal is to treat HR as a strategic lever, fine—then accept the trade-offs. You can’t demand better hiring while refusing to pay market rates. You can’t demand better performance while rewarding managers who hoard talent and burn people out. You can’t demand “data-driven” decisions and then get angry when the data reveals uncomfortable patterns.
AI can absolutely help HR deliver answers faster, reduce repetitive work, and spot issues earlier. But if companies use it mainly to scale control, speed up cuts, and dress up decisions with a fake sense of certainty, they’ll just create a more efficient version of the same distrust that already exists.
So here’s the real test: when AI makes HR more powerful inside a company, will that power be used to build stronger teams, or to enforce leadership’s preferences with better-looking paperwork?