The Next Generation of WorkTech Will Have to Show Its Work

When TIME named Cangrade to its list of America’s Top WorkTech Companies of 2026, it forced a necessary question: What does it actually mean to be part of the next generation of WorkTech?

What do we owe the HR leaders who depend on us? And more importantly, what has the industry gotten wrong that we need to fix?

One answer stands out. The first generation of WorkTech optimized for speed. The next one has to be accountable for its decisions. And most of the tools HR is currently running weren’t built with that in mind.

HR leaders are under unprecedented pressure to defend their technology. It’s no longer just about candidates asking why they were passed over or regulators checking for fair outcomes. Now, internal leadership teams are watching AI governance climb the board agenda. They want to know if the tools HR bought actually hold up to scrutiny.

Most of them don’t. That isn’t an indictment of the buyers; it’s a product problem. The WorkTech industry created this by treating explainability as an optional feature.

What Was Built vs. What Was Left Out

The first generation of AI hiring tools made a blunt trade: speed and consistency for transparency. These systems processed thousands of candidates, surfaced ranked lists, and moved pipelines faster than any manual process ever could. The efficiency was real, but it came at a cost.

What those systems rarely provided was a clear account of their logic. A candidate received a score. A hiring manager accepted it. Nobody could articulate, in plain language, what the score actually represented, which competencies were measured, how they were weighted, or whether the model had been validated against the outcomes it claimed to predict. The work happened inside the system and stayed there.

For a long time, this was the standard. The efficiency case was strong enough that buyers didn’t push on methodology. And vendors didn’t volunteer the complexity.

That has changed.

The Three Drivers of Accountability

The pressure for explainability is converging from three distinct directions.

Regulatory frameworks are moving from guidance to enforcement. New York City’s Local Law 144 requires annual independent bias audits of automated hiring tools and public disclosure of results. The EU AI Act classifies employment AI as high-risk, with requirements on explainability and human oversight. Illinois requires notice and consent before AI evaluates video interviews. These are active compliance obligations all asking the same question: can you show what your system is doing, and demonstrate that it’s doing the right thing?

Candidates are demanding trust. Candidates want to understand how they are being evaluated. Organizations that can clearly explain their processes build more trust and a stronger employer brand. Those that can’t are losing ground fast.

Internal scrutiny is intensifying. “The vendor assured us it’s valid” is no longer a sufficient answer for legal, compliance, or the C-Suite. The ability to produce documentation on what was measured, how it was validated, and what the audit history looks like is now a requirement, not a nice-to-have.

The Requirements of True Transparency

Explainability in WorkTech is not a communications problem. It can’t be fixed with better documentation or a clearer UI. It has to be built into the product from the ground up, which means the methodology has to be auditable, not just the outputs.

Three things separate WorkTech built for transparency from WorkTech that isn’t.

1. Validated measurement models
   Models must be grounded in competencies with documented evidence linking them to real job outcomes, not just proxies that correlate loosely with historical data. If a vendor can’t specify the personality or cognitive factors they measure, their score is just a number with a story attached.

2. Ongoing validation
   A single validation study from five years ago is useless. Workforce composition and job requirements change. The relationship between competencies and performance needs constant checking against real outcomes.

3. Accessible adverse impact data
   This is where vendors get quiet. HR leaders need access to data across protected categories to ensure equitable outcomes. Defensible hiring programs treat this as operational data, not something to be archived.

What this means for HR leaders evaluating tools now

The practical question is not whether to use AI in hiring decisions. That question is largely settled. The question is which tools can support the level of accountability that HR leaders are now being asked to demonstrate.

Ask every vendor these three questions:

• What exactly does your system measure, and what is the validation evidence that those constructs predict job performance?
• Has the system been audited for adverse impact, and can we see the results?
• If a candidate or regulator asks us to explain a hiring decision made with your tool, what documentation can we provide?

These are the baseline questions for responsible procurement of AI hiring software. Vendors who have built for transparency can answer them. Those who can’t are telling you exactly where they chose not to invest.

The tools that earn a place in organizations over the next decade will be the ones that treat explainability as a core requirement, not an afterthought. Building that way is harder. It requires validated methodology, ongoing discipline, and a willingness to share data that most vendors prefer to keep internal.

The industry is getting there. HR leaders shouldn’t wait for it to arrive on its own.

Gershon Goren is the Founder and CEO of Cangrade, an AI-powered hiring and talent assessment platform. Cangrade was named to TIME’s America’s Top WorkTech Companies of 2026.