Every HR leader we talk to is holding two thoughts at once: AI is moving fast, yet their organization's confidence in it isn't keeping pace. That gap isn't a lack of appetite; boards are applying pressure and nobody wants their function to fall behind. It comes down to four real concerns:

1. AI can be confidently wrong
2. Few people understand how it actually works
3. Guidance goes stale before it's signed off
4. HR can't afford to damage employee trust

None of that gets solved by moving faster. It gets solved by understanding the technology and putting the right guardrails around it.

That was the goal of two recent SHRM member webinars that MHR hosted with SHRM Atlanta and SHRM Georgia, where we looked at what's real, what's hype, and what's genuinely next when it comes to AI in HCM. This article breaks down what we covered and the questions HR leaders in the webinars were asking.

 

What's Real

AI is already creating value in HR, mostly through practical assistance, not autonomous transformation. SHRM's State of AI in HR 2026 report finds 39% of organizations have adopted AI somewhere in HR. However, of those, 56% don't formally measure whether it's working. This tells us that most have adopted a tool without actually changing existing processes.

Adoption clusters in six information-heavy areas: recruiting, onboarding, manager effectiveness, employee service, learning and development, and workforce insight. These capabilities all centre around AI reading, drafting, searching, and summarizing. That's where the technology is genuinely strongest today. We see that Adoption accelerates where these capabilities are delivered inside the tools people already use (the term for this is Assistive AI) rather than a separate destination they have to remember to visit.

Take manager check-ins as an example: MHR’s People First system pulls together prior history and goals, an AI summary of the previous check-in and AI-suggested talking points in one place where the manager reads it, edits if required and leads the actual conversation. The AI does the preparation. The manager still does the management. The same pattern shows up in other areas of the system. People First Cowork allows for Employee HR questions to be answered from trusted, permissioned company data; AI Payroll Queries explains a pay stub change and AI Job Postings drafts a job description from a brief that a person still approves. The underlying principle is: AI prepares, people decide. AI suggests, people approve. AI highlights, people interpret.

See it in action: This is what Assistive AI looks like inside People First today. Book a demo to see it for yourself.

 

What's Hype

To be clear, AI isn't overblown!. What's overblown are four specific expectations attached to it.

"AI is coming for our jobs." The World Economic Forum projects 170 million new jobs globally by 2030 (not all due to AI) against 92 million displaced, so a net gain of 78 million. Yet 52% of US workers worry about AI's impact on their job. Only 6% expect it to create more opportunity for them personally (Pew Research, 2025). The honest version is some roles shrink, most roles shift, and the function itself isn't at risk.

"Everybody else is already ahead of us." The reality is fewer than four in ten organizations have adopted AI in HR at all and of those that have, roughly a third still believe they're well behind their peers. SHRM calls this "AI FOMO." HR teams should stop chasing an imagined competitor and start moving on your own readiness instead.

"AI is transforming every part of HR." Adoption concentrates where the content and screening work is heaviest - in recruiting, HR technology, learning and development, payroll, compliance, and performance decisions have seen far more cautious adoption. That's not organizational slowness; it's appropriate caution where the stakes are higher.

"One rule will cover it." Organizational comfort with AI autonomy shifts depending on the decision AI is being allowed to make. Consider dismissing an employee in comparison to rejecting an expense claim. What actually changes is how bad it is if AI gets it wrong and whether you can undo it. That's not something a single blanket policy can capture; this is what governance is for. Governance draws the line where AI can and cannot be used.

Good governance isn't a brake; it's what makes it safe to move at all. It should define six areas:

1. Approved uses
2. Permitted data
3. When a person must approve before something happens
4. Who's accountable if it goes wrong
5. How output gets checked
6. How an employee can challenge a decision AI influenced

Not sure where your own guardrails should sit? MHR's AI consultancy can help you map all six areas above for your organization. Get in touch to start the conversation.

 

What's Next - Agentic AI

The next shift is from AI that waits to be asked to AI that works toward a goal. A chatbot waits for a question, answers, then stops. An agent receives an agreed goal, gathers information, decides its next permitted step, acts within defined limits and records what it did. It then goes round again. Agents don't replace today's assistive AI; most organizations will run both side by side.

Picture an agent with a goal to run first-round interviews. It monitors incoming results, flags a strong response against the scorecard, drafts a suggested shortlist, and routes it to the hiring manager who reviews it and decides who moves forward (this is the Human in the Loop). The agent acts only within the authority it's been given and everything it does is logged.

If agents can do more of the work, what does this mean for people? AI can handle monitoring, data gathering, drafting, and routine follow-up. People remain responsible for judgment, empathy, ethical choices and accountability. Not because the technology can't advance, but because those are human responsibilities. The goal isn't to remove people from HR. It's to focus human attention where being human matters most.

Curious where agentic AI fits in your HR roadmap? Book a demo and we'll walk through it together.

 

Your next move

MHR helps HR teams put practical, governed AI to work through People First, our AI-native HCM and payroll platform, where AI prepares and people decide. Ready to see what that looks like for your organization? Book a demo with MHR.