You already know which jobs AI will change. Do you know who's ready?

The World Economic Forum's Future of Jobs Report 2025 surveyed over 1,000 employers and found they expect 39% of workers' core skills to change by 2030. Eloundou and colleagues estimated roughly 80% of US workers could see at least 10% of their tasks affected by large language models.

Neither number tells you whether the account manager or engineer on your own team is ready for the job AI is about to change.

That's the real question for a CHRO or Head of Transformation: which of your people can already do the job AI is reshaping, and where are the gaps. Answering it takes a clear definition of what we're calling AI readiness for the changed job, then evidence that specific people can meet it.

What is AI readiness?

AI readiness is whether a specific person can do the AI-changed version of their job, to the standard the role requires, including knowing when to question or override what the AI produces.

Readiness is specific to the role, so there's no universal score, and it includes judgment, so knowing how to use a tool isn't the same as knowing when to trust it.

Why surveys and usage data don't prove AI readiness

You likely already have this data. Each source measures something real. None of it proves AI readiness.

• Self-assessment surveys
  What it can tell you: Confidence, attitudes, and perceived development needs
  What it can't tell you: Whether the person can do the work to standard
• Usage analytics
  What it can tell you: How often tools are used, and for which tasks
  What it can't tell you: Whether the output was good, or when AI should have been overridden
• Training completion
  What it can tell you: Exposure to learning content
  What it can't tell you: Whether learning transferred to the job
• AI knowledge tests
  What it can tell you: AI literacy: concepts, terms, and basic practices
  What it can't tell you: Applied performance in a specific role
• Task-based assessments
  What it can tell you: How a person performs realistic, AI-integrated work against a defined standard
  What it can't tell you: Certainty about future potential, or context the task didn't capture

EY's 2025 Work Reimagined Survey polled 15,000 employees and 1,500 employers across 29 countries and found 88% of employees use AI at work, mostly for search and summarization. Only about 5% counted as advanced users. It's self-reported, so it depends on people accurately describing their own behavior.

Usage data has a different blind spot: it measures frequency, not quality. In a study of 758 BCG consultants, Dell'Acqua and colleagues found consultants using GPT-4 did better on tasks the AI was good at. On a task outside its strengths, they were 19 percentage points less likely to get the right answer than consultants working without it. Frequent use doesn't mean good judgment.

Start with a job analysis, not a test

You can't measure something you haven't defined. Before assessing anyone, get a clear picture of what the changed job looks like.

Your hiring criteria normally come from the job as it exists today: current employees, performance data, expert input. A job being reshaped by AI doesn't have that history yet. Few people are doing it at scale, so there's no track record to draw on.

Bring together subject matter experts, people already using AI in the role, and the leaders accountable for the outcome. Map which tasks shift, where AI enters the process, which decisions stay with a person, and what good looks like. Skip this step and you're measuring something, just not something you can name.

Four questions to ask before you measure AI readiness

1. What will the job require? Document the tasks, where AI enters the work, who makes the final call, and what quality looks like, for each priority role. Update it as the tools change.
2. What evidence would convince a skeptic? Build tasks that mirror the real work, like fixing an AI-drafted client email or deciding when to overrule a model's recommendation. Score them against a rubric built on observable behavior.
3. Does the measure hold up? Check that it's consistent, that it traces back to your job analysis, and that it doesn't create unfair gaps between groups. Offer accommodations, and tell people what you're measuring and why.
4. What decision will this inform, and who owns it? Development decisions carry lower stakes than redeployment or reassignment. The higher the stakes, the more evidence and oversight you need.

Watching people work beats everything else

The best evidence that someone can do a job comes from watching them do a realistic sample of it. That's the logic behind work samples and job simulations, and it holds up: a widely cited 2022 re-analysis by Sackett, Zhang, Berry, and Lievens put the predictive accuracy of work sample tests at around .33, among the strongest of any common hiring method.

Where Vervoe fits

Vervoe is a skills validation platform built on one idea: people should be able to prove what they can do. The same approach we use to evaluate candidates through realistic, role-specific tasks works for your existing employees too, including AI readiness assessments that put people into AI-integrated work scenarios.

Every response is scored consistently, at scale, against rubrics tied to the skills that matter for the job, and every score links back to the evidence behind it. That turns "can this person do the changed work" into something you can see, while the decision stays with the people accountable for it.

If you're still working out what your AI-changed roles need, Vervoe's Skills Advisory team helps you with that upstream work: defining what matters, designing fair ways to measure it, and building a process that holds up under scrutiny.

Where to start?

1. Pick two or three job families where AI exposure and business impact are both high.
2. Run a job analysis with experts and early adopters to define the changed work and what good performance means.
3. Build or adapt task-based assessments for those roles, and pilot them with a representative group.
4. Review the results for reliability, group gaps, and employee feedback.
5. Use the first round to guide development, not big decisions, until you trust the evidence.
6. Re-measure on a set cadence. The tools and the work will keep changing.

Why this is worth the extra work

You've probably already taken the first step: knowing which jobs AI will change. The harder step is finding out, person by person, who can do the changed work and who needs help getting there. That takes a clear definition of "ready," evidence from real work, and responsible use of the results. It's more work than a heat map. It's also the difference between guessing and knowing.

See how Vervoe's AI readiness assessments can turn this into a working pilot for your priority roles.