Where employers want AI skills—and how to read the signal

What was published

SHRM released a 27-country analysis on September 22 using Lightcast active postings for IT and computer science. Its July 2025–June 2026 averages put the United States at 28.5% mentioning AI skills, followed by Canada, Argentina, Singapore, and Ireland. This describes advertised requirements in a defined occupational group, not all jobs or completed hires.

Sources: SHRM

Our analysis: turn a market signal into a work sample

For a job seeker, the useful question is what an employer expects someone to produce. An AI requirement can become an empty line on a résumé unless you connect it to a specific result. Start with a role you could plausibly apply for and identify one recurring responsibility: checking data, reviewing code, preparing a client brief, or documenting a process.

Then build a small example around that responsibility. State the task, show how you used the tool, describe how you checked the output, and record what you changed. A software developer might show a reviewed code change with tests and a discussion of failure cases. The point is to make your reasoning visible enough that another person can evaluate it.

This is our proposed response to the hiring signal, not a result tested by SHRM. It also avoids treating a national ranking as a personal forecast. A country's position cannot tell you which employer will hire you, whether an advertised requirement is essential, or how much a particular skill will add to your salary.

A practical next step

Collect ten recent vacancies for the role and location you actually want. Record the requested outcome beside each AI reference. Separate building systems from using them, and separate required experience from optional familiarity. Keep duplicate listings out of your worksheet. This exercise helps focus your learning; ten adverts are too few to support a market-wide conclusion.

Choose one repeated task and spend a week producing a demonstrable example. Ask a colleague to review the result without seeing your prompts first. Could they understand the work, detect mistakes, and continue it? Use their feedback to decide what to learn next. Keep a short record of what failed as well as what worked. Revisit the vacancy sample later, especially before committing to an expensive course or a change of role.