TL;DR

Short answer: AI native means having the judgment to know when to use AI, when to leave it alone, and how much to use, and it often means using less AI, not more. My guest, Joe Atkinson, co-founder of PromptMates, and I both reject the tool-first approach.

Key stats you need to know:

  • 44% of managers have typed employee names and performance details into public AI platforms, while only 45% of organizations have a formal written policy on that use (The Predictive Index).
  • Call center employment sits 39% below trend in the U.S., 33% below in Canada and 27% below in Germany, with entry-level workers facing the sharpest AI-driven hiring drag (Goldman Sachs).
  • 15.2% of UK people aged 18 to 24 are not in education, employment or training (ONS).

The leadership takeaway: Over the next 12 to 24 months, leaders who pair a written AI governance policy with an internal champion program will move faster than leaders who buy tools and wait for adoption to happen.

0:00
/0:46

📺 Watch the full conversation: Joe Atkinson on The Human Capitalist (YouTube)

📻 Listen on Spotify: The Human Capitalist Podcast

What Does AI Native Actually Mean?

AI native means you look at a problem and know whether AI is the right answer, and it rarely means using AI the most. Joe's clearest line from the episode was that the person topping the token maxing charts is not necessarily the AI native one.

The real test is a decision framework. Can you see a workflow and know where AI helps, where it gets in the way, and how much of it to use? Joe described community members who arrive trying to build the biggest, most complex thing they can imagine. Within a month or two they are solving smaller, simpler problems that actually get adopted by the whole team.

The counterexample is the script everyone is using. Buy an enterprise license, switch it on, and tell the team AI is now available. Joe called that approach what it is: it is not a good rollout program. Tools are a relatively small part of the work. Rebuilding workflows, writing policy, and helping people through the journey are the parts that change results.

This distinction matters for the macro picture. Goldman Sachs Research now puts AI adoption across major developed economies at roughly 15% to 20%, with hiring headwinds concentrated in a narrow set of industries and workers (Goldman Sachs). That narrowness is the point. Teams that are genuinely AI native will see gains long before the averages move.

Why Is Buying Tools Not the Same as Transformation?

Buying tools fails because transformation is an organizational problem, not a tooling problem. Joe has now run AI transformation inside a large talent function and built a community around it, and his verdict was blunt: this is not a tooling problem.

Two numbers from our current events segment show the gap. The Predictive Index surveyed 399 managers and 208 CEOs and business leaders, and found that 44% of managers admitted typing employee names and performance details into public AI tools. Only 45% of organizations had a formal written policy governing that use (The Predictive Index). Joe's guess, and mine, is that the real number is far higher than 44%, because self-reported figures understate messy behavior.

The Goldman Sachs labor market report showed the other side of the problem. Call center employment is now 39% below trend in the U.S., 33% below in Canada and 27% below in Germany, and entry-level workers face the strongest AI-driven hiring drag (Goldman Sachs). Joe and I both read the call center number with some caution, since part of that decline is offshoring and part of it is AI. The entry-level finding was the part that worried him more.

His policy point was the one I keep repeating to talent leaders. Governance done well on the front end does not slow you down. It speeds you up, because it gives people a safe space to innovate within. I spent most of my career in banking and highly regulated industries, and the lesson took me too long to learn. Get the right people in the design phase, not the approval phase. I covered a related gap in AI readiness and why it keeps widening, and my value of work matrix is where any transformation work should start.

How Should a Talent Team Actually Become AI Native?

A talent team becomes AI native by running a transformation project, not a software purchase, and the first move is finding your internal champions. Joe's playbook from the episode is worth stealing in full.

  1. Find your champions first. Every organization has people already tinkering with AI in their spare time. Give them space, freedom, and the job of figuring out what works and what does not.
  2. Build an internal council. The best versions Joe has seen run across functions. Talent team champions sit with software engineering, sales, and marketing, then bring the best practices back.
  3. Write policy before you buy more tools. If you do not have a governance policy, that is step one. Do not look at another tool or another process until you have a basic framework.
  4. Run structured education for everyone else. Top down you give champions room to innovate. Bottom up you help the rest of the team through a structured program, so nobody is left guessing.
  5. Rebuild workflows, not just tool stacks. Start from zero on each workflow and rebuild it with AI tooling in place. Access to tools is one part of that work and a small one.

Joe's warning on hiring deserves its own sentence. The people who succeed in this work are the ones willing to tinker outside their day job, and it is unrealistic to expect an entire large team to do that on top of their existing workload. That is why champions matter. They absorb the experimentation cost.

What Does This Mean for You as a Leader?

The tension in this episode is the gap between tool access and real transformation. Every leader has bought the license. Very few have rebuilt the workflow.

The leaders who win the next 12 to 24 months will pair a written policy with a champion program and then measure whether workflows actually change. The leaders who lose will keep counting seats and assuming adoption follows. I know which group I want to be in, and so does Joe.

Joe Atkinson is co-founder of PromptMates, a community and media company for recruiting and HR leaders learning to work with AI. You can find him on LinkedIn, and the community runs a free weekly live show and newsletter. If this conversation raised questions about how your function handles AI accountability, I explored the governance side in AI workforce governance and decision rights and the broader hiring picture in why HR is becoming irrelevant.

Watch the full episode on YouTube. And remember to invest in yourself, because no one is going to do it quite like you do.

Resources for HR Leaders

If you found this useful, here's what I've built for HR executives dealing with the same challenges.

FAQ

What does it actually mean to be AI native?

AI native means having an internal decision framework for when AI is the right answer to a problem, when it is the wrong answer, and how much of it to use. It is not measured by who uses the most AI tools.

Why do AI tool purchases fail?

Buying tools fails because adoption is a transformation problem, not a tooling problem. Teams need policy, guardrails, workflow redesign, and internal champions. Access to tools is a small part of that work.

How common is it for managers to put employee data into public AI tools?

The Predictive Index found that 44% of managers admitted typing employee names and performance details into public AI platforms. Only 45% of organizations had a formal written policy governing that use.

How is AI affecting entry-level hiring?

Goldman Sachs Research found AI-related hiring headwinds were strongest among entry-level workers. A 10% occupational exposure to AI cut entry-level annual headcount growth by more than 0.6 percentage points in Australia and more than 0.2 points in the U.S.

What should a talent leader do first?

Find your internal AI champions and give them the space to experiment. Write a governance policy before buying any new tool. Then run structured education for the rest of the team.

Is AI native the same as using AI constantly?

No. Joe Atkinson's point was that sometimes becoming more AI native means using less AI, not more. The judgment is knowing when not to use it.