Durable Skills are the Most Demanded, Least Developed Skills in the Economy & Has to Change

There’s a conversation happening right now in every boardroom, HR department, and education policy meeting in the country. It goes something like this: AI is changing everything. What skills do people actually need?

The answer, backed by data, is both more simple and more urgent than most organizations are prepared to act on.

Seventy-six percent of job postings request durable skills. Not technical certifications, not software proficiency. Communication, critical thinking, collaboration, adaptability, leadership, judgment. The skills that have always mattered are now the skills that matter most, and they are showing up in hiring requirements at a scale that should be getting far more attention than it is.

This isn’t a new trend. Employers have been asking for these skills for years. What AI has done is turn up the volume considerably.

What AI Actually Changes

Here’s the framing I find most useful: every time AI absorbs an execution task, what remains is more human.

When a tool can draft a report, generate a summary, or run an analysis in seconds, the value in the room shifts to the person who can evaluate that output critically, communicate its implications clearly, make a sound judgment about what to do with it, and take responsibility for the decision. Those are durable skills. And they are, by definition, the things AI cannot do for you. 

The most important story in the labor market right now isn’t that durable skills matter. It’s that the baseline expectation for human contribution is moving up the value chain faster than our education and workforce systems are adapting to meet it. Durable skills are simultaneously the most demanded and the least systematically developed skills in the economy.

The Retention Problem Nobody is Naming Correctly

Six in ten employers report having let go of a recent Gen Z hire. When you ask why, the answers are remarkably consistent: not technical gaps, but how the person communicated, collaborated, handled feedback, and showed up. These are durable skills failures, and they are expensive.

The cost of a bad hire is well documented. What’s less discussed is the root cause. Organizations are investing heavily in technical onboarding and almost nothing in the foundational human capabilities that determine whether a new hire can actually function in a team, navigate ambiguity, or grow into a role. Technical skills help people enter the workforce. Durable skills determine how far they go.

This pattern, high demand, low development, high consequence, is not a pipeline problem. It is an infrastructure problem. And it won’t fix itself.

The Definition Problem

Here’s where HR and learning leaders often get stuck: durable skills are easy to name and hard to define. Ask ten organizations what they mean by “communication” and you’ll get ten different answers, ten different rubrics, and ten incompatible approaches to assessing it. The result is fragmentation at every level: badges without validation; credentials that mean something at one institution and nothing at the next; and employers who can’t tell from a resume or transcript whether a candidate actually has the skills they need, so they fall back on degree proxies that are increasingly disconnected from what they’re actually trying to predict.

If durable skills are the currency of the modern workforce, we haven’t built a trusted banking system yet. You can’t build talent pipelines around skills that nobody defines the same way.

What Changes When Organizations Get Serious

The organizations seeing results are the ones treating durable skills as a system challenge, not a training initiative. That means starting with a shared definition. It means assessing these skills at hiring, onboarding, and throughout the employee lifecycle, not as a one-time exercise but as a continuous practice. It means building development pathways that are as deliberate and structured as technical training. And it means generating signals employers can actually trust, credentials and assessments tied to consistent definitions and performance levels.

When organizations make that investment, the downstream effects are significant: faster onboarding, stronger retention, better internal mobility, and clearer talent pipelines. More fundamentally, it changes the quality of the conversation between managers and employees, because both parties now have a shared language for what growth looks like.

The Urgency is Real

AI is not slowing down. The pace at which execution tasks are being automated means the window for building this infrastructure is shorter than most organizations realize. The skills that are hardest to automate are exactly the ones we have invested the least in defining, developing, delivering, and demonstrating at scale.

That is the gap and closing it is not a nice-to-have. It is the work we need to do to meet the moment: