20 Nov 2026

Is HR at risk with AI?

hoomph-blog-hr-at-risk-with-ai

AI in HR has become something of a double-edged sword. On one side, it’s taking the weight off stretched teams, from reducing admin to surfacing patterns in the data. On the other, it’s steadily diluting the authentic human connections that are essential to helping people – and businesses – thrive. HR teams aren’t the only ones leaning on it. Increasingly, employees are turning to chatbots to communicate their needs and frustrations too, forcing managers to decode lengthy messages packed with overly polished ‘legalese’ that take longer than ever to act upon. So, what exactly is the key to regaining balance? 

HR will always need to be guided by humans 

We’ve talked at length about the impact of AI in HR – our guide details its challenges as well as where it genuinely adds value in the modern workplace. So, we’ll be the first to reiterate that it absolutely has its place. From easing operational demands, such as CV screening and payroll processing, to strengthening strategic decision making through deep data analysis, it’s proven to provide some genuinely fruitful support for companies across the world. Not only can it interpret historical data to spot workforce trends – including flight risks and performance gaps, for example – it can also use that insight to support the development of tailored wellbeing initiatives, personalised career pathways, and even follow-up actions for team leaders. And when the regulatory landscape evolves, it can be especially helpful in summarising the headline changes before you’ve found the time to dig deeper into their meaning.

The biggest challenges of AI in HR come with overreliance. AI, by its very nature, is derivative – it can only surface responses based on the information made available to it. It doesn’t have a potted history of experience working directly with people. It can’t apply carefully honed judgment to make gut-instinct calls that affect employees’ wellbeing and livelihood. It doesn’t have the sentience to express empathy and communicate in a way that makes people feel seen and heard. It can’t pick up on the quick morning message, the forlorn look, or the out-of-character sigh you only catch by being an actual human that interacts with others. And it can’t make context-aware calls on policy changes that support compliance and the needs of their team.

What’s more, the more an employee’s experience of HR feels automated, the less they trust it. Imagine raising a personal issue – such as a bereavement, a financial concern, or a health-related challenge like menopause – and receiving a response that’s generic and hollow. People notice, and it becomes easy for rapport to thin. In fact, according to research from Conduent and Mercer, 81% of employees surveyed prefer human interaction for sensitive or complex issues. That figure is only slightly lower than in 2023, which suggests that years of normalised chatbots and self-service portals haven’t dented trust in human judgment. When leaders choose to route conversations through AI, concerns that would once have been quickly and quietly addressed can start surfacing as absences or formal complaints. And that’s a very expensive way to save a few hours. 

How employees are using AI to up the ante

On the flip side, employees are increasingly using generative AI to research workplace policies and regulations, draft grievances, prepare for disciplinary meetings, challenge redundancy processes, and plan for tribunal. It’s not hard to see why. People who feel anxious, unheard, or unsure of their rights will reach for whatever gives them confidence and clarity at 11pm on a Sunday. And often, it feels like the only place to ask a ‘silly’ question without judgment.

But it’s landing on HR teams hard. According to research from Brightmine, nearly nine in 10 organisations (87%) have seen a rise in AI-assisted grievances. The challenge isn’t necessarily the act of consulting technology, but the undisputed trust in its outputs. AI doesn’t always verify the facts. It can quote the wrong policy, mangle an employment right, or exaggerate details that weren’t in the original complaint. Its answers are also often optimistic, designed to please the user, resulting in inflated expectations. Not to mention, they bring volume over clarity. AI-written grievances tend to be lengthy and highly legalistic, and can look impressive while containing irrelevant or inaccurate content that only requires additional time to untangle and understand. 

Yet, at the end of it all, there’s still a sensitive issue to manage. The information may be badly packaged, but the feeling underneath is still real. So, the question isn’t necessarily why employees are using AI, but how to make them feel comfortable enough to talk to their line manager first – and directly.


What our human HR experts say

The disadvantages of AI in HR impact both sides of the relationship, but line managers are well placed to help regain the balance and sever the damaging AI loop that’s eroding crucial, human workplace connections. Nobody sits closer to the day to day. By understanding the team and its culture, these individuals can build genuine psychological safety that lets people know they don’t need to rehearse, polish, or outsource what they want to say – they can just say it.

Here’s how line managers can lead with connection:

  • Make talking the default. Hold regular 1:1s that aren’t just task updates. Ask open questions like “What’s getting in your way?” and then wait for the answer.

  • Respond well to the small stuff. How you react to a minor worry tells people whether it’s safe to raise a bigger one. Thank them, be curious, and follow up.

  • Say it out loud. Tell your team you’d always rather hear about a concern when it’s small and messy than polished and escalated. 

  • Be comfortable not knowing. Saying “I’m not sure, let me find out and come back to you by Thursday” builds far more trust than an overconfident guess. 

  • Make the process easy. If people can’t tell how to raise something or what happens next, they’ll ask a chatbot. Explain it in plain English and signpost where to go.

  • Choose people over paperwork. If an AI-written complaint lands on your desk, don’t lead with suspicion or defence. Acknowledge it, pick up the phone or book time with them, and ask them to tell you about their issue in their own words

  • Be patient with change. People are habitual characters, and if they’re used to constantly defaulting to a chatbot, that won’t shift overnight. Meet it with calm and kindness rather than frustration. Similarly, every time someone chooses to talk to you directly instead, a gentle “I’m glad you came to me” goes a long way.

Will AI replace HR jobs in future?

Parts of the admin, quite possibly. But the trust, judgment, and human connection that help people thrive aren’t going anywhere, and that’s exactly what we help businesses build. Our line manager training gives managers the skills and confidence to hold difficult conversations, build psychological safety, and become the person their team turns to first. Get in touch – we’re here to help.

SHARE THIS ON: