Let the bots crunch; let the people decide.
“At this point, so many people are talking about AI, we might need to do a background check on AI.” That’s how one attendee greeted Andy Gallion, InCheck’s co-founder and Chief Development Officer, and it stuck because it’s true! SIA 2025 showed us, over and over again: the fastest teams automate the busywork and double down on human judgment where it counts.
So if you left SIA last week thinking “AI will replace recruiters,” we heard something a little different. Yes, the tech certainly helps, but the make-or-break moments are still human. Here’s what we heard in the hallways, and how top teams are blending speed with human engagement.
Takeaways from Andy Gallion, InCheck Co-Founder and Chief Development Officer
1) Tech pulls its weight, but it shouldn’t run the show
The vibe was: “AI is great at the repetitive work. But don’t hand it the keys.” Resume parsing, scheduling, and credential checks may be humming with AI assistance. But when the stakes are human safety and bedside chemistry, it’s still important to default to human judgment. Not because we’re sentimental. It’s because that’s where the risk lives. If the model lands you with a culture mismatch or misses a red flag in the story behind a gap, you feel it in patient care, not just in a KPI.
We see this clearly in Occupational Health Services: when teams over-automate immunization validation, titer interpretation, and fittesting checks, they invite rework, delays, and compliance misses. Instead, a human shepherd can clarify ambiguous results, catch jurisdictional nuances, and coach candidates through the clinical bits, improving accuracy, auditreadiness, and the candidate experience.
My takeaway: Tools are power steering, not autopilot. You still have to drive.
2) The outcome hinges on a few human moments
What I heard in the hallway: “My north star is: would I let this person care for someone I love?” A machine won’t tell you that.
Whether it’s in a final interview, a motivation check-in, or expectationsetting with clients. Those are the leverage points. You can learn more in ten minutes of honest backandforth than in fifty screens and a recommendation score. People can sense, “Is this person curious? coachable? steady under stress?” The machine can’t.
Service will always beat speed when the stakes are high. A bot can book an appointment, but it takes a seasoned professional to disarm a nervous candidate, de-escalate an issue, or spot a risk hidden between the lines. Those small human moments are where fall-offs shrink and placements stick.
3) Bias is like glitter. It gets everywhere unless you plan for it.
Everyone likes that AI levels the playing field on speed. But we saw the usual traps: nonlinear careers that get penalized, name formats that trip filters, gaps that hide caregiving or international moves. We see that the shops that will remain ahead of the curve are building human review as a standard step, not a rescue. “This person doesn’t pass the autoscore. Here’s why we’re advancing anyway.” That little note can save you in audits and, let’s be honest, find you great hires.
My opinion: Fairness isn’t a feature; it’s a habit you practice.
Observations from Joe Doyle, VP of Sales & Account Management
1) Candidate experience is won in early interactions
Candidates will use your portal and reply to automated nudges. But what lowers stress (and dropoff) is a human who says, “Here’s the next step, here’s the ‘gotcha’ to avoid, and text me if something weird pops up.” Early human touch despooks the process. People want to feel seen, not just scanned.
Simple fixes that support a smoother process:
- Messages from a real name, not “noreply.”
- A short expectation-setting call early.
- Clear “what happens next” in plain language across messages, updates, and platforms.
- A live contact for Occupational Health questions—before the lab visit—not after someone fails to complete the screening.
- If things stall, “More alerts” is not a strategy. Fix the funnel so humans handle the ones that actually matter.
2) The “Faster, Cheaper, Better” Trap
When the demand is “faster, cheaper, and better,” the tempting move is to trim human checkpoints and let automation run hotter. But if faster and better are truly the goal, the honest follow-up is: are you willing to invest more? In clinical settings, especially, removing steps and experts doesn’t simplify the process; it likely shifts cost into rework, delayed starts, and even reputational damage.
The advanced takeaway from SIA was this: if you want faster and better, that usually means more of something. More expert review, more contextgathering, more proactive communication. So the question to ask isn’t, “Who’s cheapest?” It’s, “What are the true requirements, where does human judgment matter in our funnel, and are we over-automating those spots?”
In Conclusion
Talent acquisition leaders aren’t stuck on whether to use AI. They’re stuck on how to prove it helps without faking precision.
The reality check is straightforward. Speed is table stakes. Trust is what wins deals and keeps the team intact. Compliance is the guardrail that keeps you out of the ditch. And great screening partners—the ones who listen well and ask real questions—are your edge, not an expense. If your model forgets that, you’ll feel it in drop-offs, rework, and churn.
So what do you do next?
- Install a review loop for edge cases and log where and why you do, or do not, advance candidates. Those notes pay off in fairness and audits.
- Make the decision lines explicit: write down what the tool does versus what a person decides, and stick to it.
- Keep Occupational Health humanled to protect accuracy, compliance, and candidate experience.
- When someone asks for “faster, cheaper, better,” clarify which two they actually mean—and what they’re willing to invest in for “better.”
Bottom line: Put people in the foreground where nuance, safety, and trust live. That’s not antitech or antiprogress. That’s prooutcomes. Speed is easy to buy; judgment is what compounds.