The Real AI Threat

Journalists in the legacy media haven't been this excited about the end of the world since Y2K. Now they can barely contain themselves as even titans of the technology industry warn that artificial intelligence may eventually destroy humanity.

I'm not about to place any bets on that happening. My suspicion is that much of the hysteria will eventually fade, just as the predictions of worldwide catastrophe did when the calendar turned to January 1, 2000.

But there is a very real AI problem unfolding right now, and it's one that isn't getting nearly enough attention.

AI is making it increasingly difficult to know who you're dealing with.

At Paradigm, we're seeing this firsthand in recruiting, particularly for technical positions.

Within minutes of posting a job, we can receive dozens, sometimes hundreds, of résumés. Many look terrific. The candidates have precisely the skills we're looking for, impressive employment histories, and experience with every technology mentioned in the job description.

There's just one problem. Some of these candidates aren't real. Others may be real people using fabricated qualifications or stolen identities. Still others may have the necessary skills but aren't located where they claim to be.

And AI has made the entire process remarkably easy.

A résumé can be tailored to a job description in seconds. A convincing employment history can be manufactured. Even interview answers can be generated with AI assistance. What once required considerable effort can now be accomplished almost instantly and on an industrial scale.

We've encountered some peculiar situations. During interviews, we've heard what sounds like a call center in the background, with phones ringing and other conversations taking place. We've received two résumés from candidates supposedly living in the same small Florida town, both applying for a highly specialized technical position. Coincidence? Perhaps. But it certainly got our attention.

The problem becomes particularly troublesome with remote positions, where an employer may never meet the employee in person.

Our recruiter, Meg, has become something of a detective.

She looks for inconsistencies in résumés and LinkedIn profiles. A profile with no photograph, no posts, very few connections, and only a handful of jobs can raise questions. None of these things proves anything, of course, but taken together with other inconsistencies, they may warrant a closer look.

In fact, Meg has become so frustrated with questionable applications that she no longer wants to advertise technical positions on free job boards. Instead, she prefers actively identifying and approaching candidates through professional recruiting tools.

And then there's the interview.

We've learned that asking conventional technical questions isn't always enough. AI can provide excellent answers to almost any question, sometimes while the interview is taking place.

So we try to make interviews more conversational. Ask candidates to explain how they solved a particular problem. Follow up with questions about the decisions they made, the mistakes they encountered, and what they might do differently. Ask about something unrelated to work. Get them talking naturally, rather than simply reciting prepared answers.

The objective isn't to trick legitimate candidates. It's to establish that we're talking to a real person with genuine experience who can think independently.

There are also serious consequences beyond wasting a recruiter's time. A company that unknowingly hires someone using a false identity could expose confidential information, customer data, intellectual property, and internal systems to people who should never have access to them.

The irony is that AI, a technology designed to make us more productive, is forcing us to spend more time verifying things we once took for granted.

I don't know whether artificial intelligence will someday become smart enough to take over the world. But I do know that it's already making it easier for people to deceive one another.

At Paradigm, we've always believed that recruiting is fundamentally about people, not just résumés and keywords. Technology can make us more efficient, and we intend to use it. But it cannot replace human judgment, personal attention, and relationships.

Perhaps that's the real lesson here.

The more artificial our world becomes, the more important the human element becomes.