Companies spent years telling workers they needed degrees, certifications, and endless experience to get hired. Now, apparently, knowing how to casually throw around AI terminology in a job interview might be enough to land a position.
That’s the vibe coming from new research released by TestGorilla, which says 53 percent of hiring managers now value AI fluency more than deep subject matter expertise. At the same time, 59 percent of organizations admit they made a “bad AI hire” within the past year.
Honestly, I believe it.
We are living through a strange moment where everybody suddenly claims to be an AI expert. Scroll LinkedIn for five minutes and you’ll find people calling themselves AI strategists, prompt engineers, workflow architects, and automation consultants. Some of these folks absolutely know their stuff. Others? Not so much.
According to the report, many candidates are learning the language of AI without necessarily understanding how to apply it in real-world situations. Terms like RAG, prompt chaining, vector databases, and agentic workflows can make somebody sound incredibly smart during an interview. But saying the words and actually building something useful are two very different things.
The study surveyed nearly 2,000 hiring leaders across the US and UK, and one thing became pretty clear: a lot of businesses are still figuring this out as they go.
TestGorilla says some companies consider simple awareness of AI tools to be enough qualification for a role. Others leave AI evaluations entirely up to individual hiring managers, which sounds like a recipe for inconsistency. One manager might want a true technical expert while another just wants somebody who sounds confident talking about ChatGPT and automation.
That creates what is basically a corporate vibe check.
And frankly, some executives probably do not fully understand AI either, which makes the situation even messier. If the person conducting the interview barely understands the technology, how exactly are they supposed to separate genuine expertise from buzzword soup?
One of the more interesting details in the report is the difference between the US and UK. American companies reported far more frequent AI-related workplace errors than organizations in Britain. TestGorilla suggests UK employers may be doing a better job defining what AI fluency actually means before hiring around it.
There’s also a bigger issue hiding underneath all this hype. AI tools can absolutely make workers more productive, but they are not magic. Somebody still needs actual expertise. A chatbot cannot replace years of experience in software development, finance, healthcare, journalism, or engineering. At least not yet.
The danger is that companies become so obsessed with appearing AI-first that they accidentally prioritize confidence over competence. And once that happens, you end up hiring people who know how to talk about AI instead of people who know how to use it effectively.
Folks, those are not the same thing.
