AI in recruitment: from hype to hard choices

Artificial intelligence is reshaping recruitment far beyond chatbots and resume parsing. For Adam Godson, former CEO of Paradox and Head of Workday’s talent acquisition products following Workday’s billion dollar acquisition of Paradox, the real story isn’t the technology itself—it’s the decisions, governance, and change management that surround it. “Everyone thought it was the software that was hard to do,” he reflects. “It’s actually the decisions that are hard to make—the decisions that it forces you into.”

In a wide ranging conversation at Workday’s Dublin office with Enterprise Times, Godson outlined how AI is transforming hiring. What good governance looks like, why change fatigue is a real risk, and how enterprises can move beyond pilots to meaningful impact.

Governance: bias, stability and explainability

As enterprises accelerate their use of AI in recruitment—whether for screening, scoring CVs, or managing candidate conversations—governance is moving from a theoretical concern to a practical necessity.

For Godson, rigorous evaluation is the starting point, “A couple things that are happening for us is standardised ways to do evaluations inside of agents, to understand what things are working, when they’re not, when they’re not working as expected.”

He highlights several core dimensions of governance:

• Bias testing – “Anytime we’re doing a type of scoring of CVs… being able to have the methodology to do bias testing, stability, similarity.”
• Stability – “As we run this 100 times, do you get the same things other times? Those types of things to understand what type of variance we have in AI models.”
• Explainability – which he calls “one of the tricky things,” especially with modern large language models. Explainability is the ability of an AI system to clearly show why it produced a particular output or decision, in terms that humans can understand and verify. In this context, it means being able to give a reliable, grounded reason for actions like scoring a CV or matching a candidate, rather than just a vague or post hoc story.

Beyond model metrics, Workday also focuses heavily on conversation quality. The company employs a team of data taggers and uses statistical sampling to review real interactions:
“We judge conversations as part of our methods to understand what people are asking… Are they getting what they needed? Is that a high-quality response?”

Ironically, while AI is the subject of this governance, it is also part of the solution: Workday increasingly uses LLMs as evaluators in their own right.

The future of recruiters: less admin, more convincing

Headlines often warn of the “death of the recruiter”, but Godson rejects that narrative. Recruiters will change—but they will not disappear.

“The administrative tasks go away, but that was never really the job. That was the by-product of the job.” One of Workday’s most common use cases is interview scheduling at scale. The impact on operations is immediate and measurable, “We see regularly managers say, we’re going from having a room of 30 interview schedulers to having 15 or 10.

“As long as there’s competition for talent, there will be recruiting. Look at the military, you look at anywhere there is competition for the best talent… recruiting will continue to matter. But it’s not the administrative type of clicking the button recruiting. It’s the ‘I’m convincing you that this is what you should do with your life.’”

In other words, AI strips away the low value tasks to reveal the core of the profession: persuasion, relationship building, and judgment.

Skill shortages: AI as both problem and solution

In recent years, enterprises have struggled to hire data scientists, data engineers, and AI specialists. Has AI itself solved that shortage? “It’s both the problem and the solution,” says Godson.

He distinguishes between two very different kinds of AI skills:

1. Foundational model researchers – “You see a shortage, of course, at the foundational model company, but those are relatively few companies—that’s Anthropic, Google, Meta, OpenAI… typically PhD level researchers.”
2. Applied AI talent – which is where he believes the real shortage lies. “The real shortage we see is people to apply AI change knowledge in a company. A company wants business transformation. How do I do it?”

Enterprises don’t just need people who can build models. They need people who can embed AI into processes, workflows, and experiences that change business outcomes.

Why Workday is betting on in person AI innovation

Workday is investing heavily in AI, including a new Innovation Centre near Trinity College Dublin. In an era where many technology companies are still debating remote vs. hybrid work, Godson sees in person collaboration as a competitive advantage—especially for AI.

“We think being in-person, and getting lots of molecules bouncing off each other, leads to better innovation. We have a hypothesis that AI driven development can happen even faster that way.”

Coding itself, he argues, has changed, “The task of coding is just different than it used to be… And I think that’s going to lead to in person teams.” He also suggests AI is shifting long standing assumptions about global talent strategies. “For the last 50 years, companies, broadly speaking, have been chasing geographic arbitrage in some fashion. AI presents a new frontier where that may not be the case.”

From inputs to outputs: rethinking talent management

The combination of remote work and AI is forcing a fundamental rethink of how organisations measure performance. During COVID, many firms claimed to focus on outputs rather than inputs (hours, physical presence, visible activity). Godson believes AI will finally force them to follow through. “With AI, you’re going to see a massive variance in the output of some employees. Some of those employees are going to be 10x what others have done.”

The implication is profound.

“The transformation with AI will force companies to focus more on outputs than ever before. This will lead to a massive distribution of productivity. As a result, companies are going to want to do a massive redistribution of rewards. Furthermore, reward the best people the most.” Talent management, in this vision, becomes far more performance and outcome driven, with AI both enabling and exposing those differences.

Why Big Bang AI transformations rarely work

Enterprises know they “need to do something” with AI, but many are stuck between hype and hesitation. Businesses are unsure whether to go all in or proceed cautiously with pilots.
Godson doesn’t believe in vague “all in” pronouncements. “I still don’t know what it fully means for an organisation to say they’re going, all in on something. How many transformation projects could you run at a given time?”

Instead, he advocates a disciplined, problem driven approach. “At this pace of change, it is prudent to get some stuff all the way done. What I tell people is, first start where it hurts, and then find something where you can be successful. Go win one. And then do it again and again, and those transformation projects will knock down.”

The trap, he warns, is endless experimentation with no integration. “If it’s not properly integrated back into the core of the business, it’s just seen as an experiment and a pilot. Then people never really take it on.”

The real constraint: human capacity for change

For all the talk of AI’s exponential progress, Godson believes the limiting factor isn’t the technology. “The actual barrier to everything is humans—how fast can you change?” He lived this during Paradox’s hyper growth phase. “We were tripling in size every year. At one point I calculated that our median employee had three months of experience. It was a knowledge problem… The same is true with change.

“How fast people can absorb and adopt change is, frankly, the limiting factor in how fast AI can change work.”

That leads to one of the few things that does keep him up at night. “I worry about change fatigue in people, and I worry about being sure that we can deliver on some of the possibilities that AI presents.” Yet he remains optimistic—especially about recruitment.

“Recruitment is changing fast, and there are possibilities that have never been there before. But the core is the same. In many ways it’s a back to basic. AI is a chance to get people connected with people again, and to do some good in the world.”

The post AI in recruitment: from hype to hard choices appeared first on Enterprise Times.