The AI Reality Check: Challenges We're Actually Facing in 2026

Remember when AI was going to solve all our people problems? Yeeeaaaaeah…  About that...

We're now two years past the initial AI explosion in HR and talent engagement, and whilst I'm not about to declare it all a failure (far from it), we're definitely in the 'messy middle' phase. The shine has worn off, and we're left dealing with some properly challenging realities.

Numero uno - The authenticity problem.

Increasingly, people can tell when they're interacting with AI, and they're making judgements about you and/or you organisation based on it.

Those LinkedIn posts you’re spitting out that are 100%, largely unchecked AI guff (even this one uses AI a bit but I lead, it refines). That chatbot answering candidate questions poorly at 2am with the personality of a dead slug? Those dire AI brain burps you’re leaving in peoples’ comments section. Brilliant in theory. In practice, it's giving canned responses that make you and your organisation feel about as genuine and warm as a self-checkout machine telling you there's an 'unexpected item in the bagging area!'

Em dashes as indicators of AI over usage are so H1 2025. Now its the, “but here’s the hard truth...” ers. The, “this is what no one is telling you...” junkies. The “So, what’s your take?” peddlers. We see you. We roll our eyes at you, and that, “be authentic” dross you’re ramming down our throats is sticking in yours like a fish bone grabbing on with velcro fingers - ugly cat-heaving-face heaving na’ll.

I’m seeing too many people and organisations optimising for efficiency at the expense of connection. And in a world where human relationships and candidate experience are increasingly everything, that's a real problem.

Rightio. With my soapbox fully warmed up, let’s move along to number two - the bias we didn't eliminate.

We told ourselves AI would remove human bias from recruitment. Instead, we've just automated it at scale. Your AI screening tool is still filtering out career-breakers who took time out, still penalising unconventional career paths, still favouring candidates who've learned to game the keyword system.

What we were told we’d remove is now happening faster, and with a veneer of objectivity that makes it harder to challenge. I urge you to really challenge any tech vendors claiming to be able to do this for you. Get your Legal, Data and Compliance teams all over them. Check every nook and cranny of their SaaS. Yeah, it might slow down the procurement process but better that then a hefty lawsuit by someone who decides to challenge their rejection and finds out there was no human interaction on the road to making it.

What next? Ah yes - the commodification trap. When everyone's using the same AI tools - and let's be honest, there are about five main players - how exactly are you differentiating your employee experience? Your AI-generated job descriptions sound like everyone else's if you're not feeding them with robust guidance about your culture and brand identity. Your automated scheduling feels identical to your competitors'. Your chatbot uses the same training data. We're accidentally creating a talent market where everything feels homogenised. And rule number #1 of the best brands? Be distinctive! It’s starting to looking increasingly like a number of people didn’t get that memo. Although, let’s be honest, that was a problem in this space way before AI came along.

“So, Mr. Moany Moanerson . how do you suggest we move forward?” I hear some of you grumble. And rightly so. It’s easy to throw rocks from the sidelines and I don’t want to be "that guy”.  I'm not suggesting we bin AI and go back to manual everything. That's not realistic, and frankly, some applications genuinely work brilliantly. But we need to get more intentional. We need to ask harder questions about where AI adds genuine value versus where it's eroding the human elements that actually matter. We need to design systems that augment human judgement rather than replace it.

The organisations I see succeeding aren't the ones with the most sophisticated AI. They're the ones who've figured out the uniquely human aspects of their employee experience and protected those fiercely whilst using AI for the genuinely mundane stuff.

Maybe the real challenge of AI in 2026 isn't technical at all. Maybe it's remembering what makes us irreplaceably human - and designing our AI systems accordingly.