CAREER TRANSITION

By Ric @ Jobric · August 2026 · 6 min read

Here is where hiring quietly landed in 2026. You used an AI tool to help write your application. On the other end, an AI read it, and increasingly, an AI ran or scored your interview. Two machines, talking to each other, with two humans sitting behind them hoping it works out.

That is not automatically a disaster. Automation on both sides could be fine. The problem is smaller and sharper than the panic suggests: nobody agreed to the rules, and most of the time, only one side knew the machine was in the room.

Let me read this the way I read everything here. Without the hype.

The numbers are real, and they moved fast.

Roughly 74% of US job seekers now use AI tools somewhere in their search. And 63% report facing an AI interview, up 13 points in just six months, according to Greenhouse's 2026 Candidate AI Interview Report. Greenhouse's own CEO, Daniel Chait, has called this the "AI doom loop": your bot writes, their bot reads, their bot interviews, and somewhere in there a human decision is supposed to happen.

I am not going to tell you the machines are evil. They are not. The absurdity is not the technology. It is that a two-sided system got built with no shared agreement about how it works, who gets told, or what happens when it is wrong.

This is the part worth slowing down on.

Read those together. The issue is not that a machine is in the room. It is that only one party knew it was there. That is a consent problem, not a technophobia problem. When one side gets to automate silently and the other side finds out later, that is not efficiency. That is an information imbalance dressed up as progress.

And people are voting with their feet. More than a third have quit a process over it. That is not a fringe reaction anymore.

Here is where it gets personal, because a title-matching filter is a very literal reader.

It sees "Staff Nurse." It does not see runs a floor under pressure, coordinates twelve people, catches the error before it reaches the patient. It sees "Accountant." It does not see risk instincts, controls thinking, the ability to smell a number that is lying.

The pivot is the one thing these systems are worst at reading. They match on your last title, not on what you can actually do. So the strong candidate who is deliberately changing direction reads as a false negative: screened out before a single human ever weighs the real match.

If you have felt this, you are not imagining it. This is the specific, mechanical wound of a career change in 2026. You are being matched on where you have been, by a system that has no idea where you are going, and no instruction to ask.

Now let me push back on my own alarm, because the fear gets sold in inflated numbers and I am not going to do that to you.

You have probably seen claims that 83% to 99% of employers screen with AI. Those numbers are vendor-flavored and overstate reality. The neutral data is more modest and, honestly, more useful.

SHRM finds that 27% of organizations use AI specifically in recruiting, and 39% currently have AI adopted across HR overall, with another 7% planning to launch this year, putting total expected 2026 adoption at around 46%.

So read the direction, not the denominator. Adoption is real. It is growing fast. It is also uneven and nowhere near universal. Most employers are not running a fully automated gauntlet. A meaningful and rising share are using AI somewhere in the process, often without telling you.

That is the true version. It is less cinematic than "the robots reject 99% of you," and it is the one that lets you make good decisions instead of panicked ones. Refusing to inflate the number is the point. The trust is the product.

One more piece of honesty while we are here: you will read that AI is gutting entry-level hiring. Some of that data shows correlation, not clean causation, and serious researchers, including a recent LSE-affiliated paper (Lambert and Schindler, "The Broken Ladder"), contest the tidy story, arguing remote work explains more of the decline than AI does. Treat "AI took the jobs" as a claim under argument, not a settled fact.

No lectures. Four things.

Here is the thing that bothers me most about a silent title-match rejection: it never tells you why. It just goes quiet.

That is the exact problem we built against. AI Fit Analysis reads your skills and your evidence, not only your last title, and then it shows you what aligns, what is a stretch, and why a role made the cut. It is the opposite of a silent no. And Focus Lens lets you hold more than one direction at once, so your matches reflect where you are going, not only where you have been.

The market runs on opaque machines. Fine. We made the machine that works for you explain itself. That should have been the default.

The surveys count interviews and tools. They are good at that.

They do not count the person who quietly stopped applying because a bot rejection felt exactly like a human one. They do not measure the good pivots that never happened because no filter could read them. They show adoption. They do not show consent, dignity, or the real cost of being misread when you were, in fact, the right person.

Numbers are honest about scale and silent about everything that makes this hurt. Hold both.

The transparency question is moving from etiquette to policy. When 57% of candidates want disclosure legally required, that is a signal regulators tend to notice eventually. Expect "did you know an AI was scoring you?" to stop being a courtesy and start being a rule.

Until then, the burden is unfairly on you to be legible to a machine that will not announce itself. Write like it is reading. Ask when you can. And do not let a literal-minded filter convince you that a pivot is a flaw.

That's the update. Now go do something that isn't job searching.

Ric @ Jobric