Every HR and Talent Acquisition conference right now seems to give roughly the same messaging – AI is transforming recruiting. AI is revolutionising HR. AI is everywhere, doing everything, and the organisations that haven’t fully embraced it are already falling behind.
I recently interviewed my friend, co-author, and podcast host Matt Alder, on the HR Means Business podcast as he has been speaking to practitioners, vendors and thought leaders about the future of talent for over a decade. His honest assessment of where we actually are is rather different from the narrative being sold.
“It’s a very mixed picture,” he told me “There are certain sectors where it’s full speed ahead – principally front-line volume hiring – but the rest of the talent space is still largely at the pilot and experimentation stage. The bulk of organisations haven’t moved beyond the most simple use cases”
Interview scheduling. Rewriting job descriptions with an LLM. Things that, as Matt points out, aren’t really AI at all in any meaningful sense.
The Jagged Frontier Problem
Matt borrowed a phrase from Professor Ethan Mollick to describe where AI actually is right now: a jagged frontier. Remarkable at some things. Surprisingly poor at others. And often dangerously confident about the things it gets wrong.
The organisations making real progress, he argues, share a common starting point – they began with the problem, and not the technology. They find out what was broken in their process first – and then worked out how AI could help fix it.
The companies that simply plugged AI tools into existing processes found that speeding up something that wasn’t working just made it worse faster.
This distinction matters enormously. And it leads directly to Matt’s framework for what AI readiness in talent acquisition actually requires: five things that need to work together rather than in isolation.
Process architecture – genuinely rethinking how recruiting works, not just automating the current version.
Decision design – being explicit about where human judgment sits in the process and why.
Team capability – understanding what skills recruiters actually need in an AI-enabled world, which involves more bottom-of-funnel relationship work and less top-of-funnel screening.
Data and measurement – moving beyond the metrics that were easy to capture to the ones that actually matter to the business.
Governance and trust – both the regulatory dimension and the equally important question of whether candidates trust the process they’re going through.
“People talk about governance all the time,” Matt observed. “Data and measurement get discussed occasionally. Capability, decision design and process architecture – hardly at all.”
The Candidate Side Nobody is Talking About
For Matt, the most underreported story in recruiting right now isn’t what employers are doing with AI – it’s what candidates are doing.
Application volumes are surging. Resumes are being auto-generated and customised at scale. And while the industry conversation frames this almost entirely in terms of cheating or gaming the system, Matt sees it differently.
“Candidates are just using the tools available to them. The CV is the wrong format for the age we’re living in – and candidate AI is exposing that.”
The real innovation in talent acquisition, he believes, is going to be driven from the candidate side, not the employer side. Candidates move faster. They adopt tools faster. And the organisations responding intelligently – moving skills assessment earlier in the process, designing experiences that are genuinely agent-resistant – are the ones that will pull ahead.
The Ownership Vacuum
Focus on this picture for a moment. Most organisations have no clear owner for AI strategy. In some it sits with technology. In others, legal or compliance. In many, nobody owns it at all – it’s just noise. In very few organisations does HR or Talent Acquisition take the lead.
The consequences of that vacuum are significant. Employees are using AI for work without telling their employers – and the regulatory and bias risks that creates are substantial. Simply banning AI use, as some organisations have tried, doesn’t work. People use it anyway, just less transparently.
“The risks of doing nothing, or of banning it, are enormous,” Matt says. “Neither of those things reflects the reality of what’s actually happening.”
What Good Leadership Looks Like Here
It comes down to two things. A clear vision – not just for the technology, but for what you actually want TA or HR to achieve, and how AI fits into that journey.
And psychological safety – creating an environment where people can talk honestly about how they’re using these tools, where pilots are allowed to fail without blame, and where the organisation learns collectively rather than hiding its experiments.
The technology is moving faster than any previous wave of change. But the fundamentals haven’t changed at all.
- Start with the problem
- Design the process
- Build the capability
- Be honest about where you are
The organisations that do those four things will look very different in 18 months time from the ones still debating whether AI is ready for them.
AI IS ready for them. The question is whether they’re ready for AI?
Check out the full podcast conversation here – https://www.hrhappyhour.net/episodes/making-ai-matter-leadership-culture-and-the-future-of-hiring/ or through the image below
