Ask most managers how their team works today, and you’ll hear about AI tools drafting reports, summarizing calls, and cleaning up data. Then, pull the job description for those same roles. Odds are it reads the same as it did two years ago.
A new report from Oyster, a global employment platform, puts numbers to that gap. In its 2026 Global AI Talent Report, 95% of HR professionals said employees at their company use AI tools daily in roles where AI isn’t part of the official job description. Two-thirds have added AI skill requirements to non-technical roles like marketing, HR, and operations. Of those, 46% never formally told employees about the change. And 62% of companies that added AI expectations to existing roles changed neither the title nor the pay. Oyster’s CEO, Hadi Moussa, says we’re missing a “language” for describing what people can do. Titles, credentials, and years of experience used to carry that information. They carry less of it every month.
LinkedIn sees the same shift from the candidate side. At its Talent Connect conference in September, CEO Dan Shapero said the skills employers hire for are changing at twice the rate of the skills candidates list on their profiles. His summary: “Every job is going to be a new job.”
What the gap does to hiring. When the job description lags, every step after it lags too. Postings attract people who match the old version of the role, and screening rewards the old skills. Oyster found 68% of companies ask candidates what they have built with AI, yet 53% say the answer doesn’t affect hiring or pay decisions. We’re asking a good question and then setting the answer aside.
What it does to retention. Your people notice. Only 38% of HR leaders said an employee who automated part of their own job would be rewarded, and 24% said that employee would be handed more work. If learning a new tool earns a heavier workload with the same title and pay, your most adaptable people will conclude the reward is somewhere else.
Where to start. Pick your three highest-volume roles. Sit down with a manager and two people who do the work, ask how the job gets done today, and rewrite the description from that conversation. Separate the skills someone needs on day one from the ones you’ll teach, and say so in the posting so strong candidates missing one tool still apply. Decide what the change means for title and pay before you announce it, then announce it. Employees will fill silence with their own answer.
AI can help you draft a clean job description in seconds. What it can’t tell you is how the work changed on your team. That comes from talking with the people doing it.
The job description is the first thing a candidate reads and the yardstick a new hire measures the job against. If it describes a job that no longer exists, it’s worth an afternoon to fix.
Sources: Tom Starner, As AI upends job descriptions, it’s time for “replacement language,” HR Executive, Oct. 23, 2026 (Oyster 2026 Global AI Talent Report). Jen Colletta, LinkedIn CEO: Every job is going to be a new job, HR Executive, Sept. 30, 2026.