AI is everywhere right now, bringing new tools, bold promises, and mounting pressure to “do something” quickly. But for HR leaders, the real challenge isn’t the technology itself. It’s deciding where AI actually helps, where human judgment can’t be replaced, and how to move forward responsibly.
In a recent webinar, Mira Greenland, CRO at INTOO, sat down with Trent Cotton, Head of Talent Insights & Analyst Relations at iCIMS, to cut through the noise and focus on practical, leadership-driven guidance that addresses HR’s particular concerns and needs.
The goal was to help leaders confidently answer three critical questions:
- How do we distinguish high-impact HR use cases for AI from areas where human judgment is irreplaceable?
- How do we avoid “shiny object” investments that feel innovative but don’t move business outcomes?
- How do we protect company and personal data as we explore and scale AI solutions?
The Current State of AI Adoption in HR: Lots of Use, Limited ROI
One of the challenging realities of current AI adoption is that its usage is rising faster than its measurable value.
Trent pointed to a growing gap between executive expectations and operational readiness. Many leaders assume AI is a “magic wand,” but the truth is more grounded:
- AI outputs are only as good as the data and systems behind them.
- Many organizations still have fragmented, inconsistent, or poorly governed data.
- AI requires ongoing oversight; not “set it and forget it.”
As Trent put it, AI still needs to be managed like anything else that impacts your business.
“Inspect What You Expect”: The Most Important Mindset Shift
Trent returned often to a simple but powerful principle:
You have to inspect what you expect.
In other words, you don’t implement AI and walk away. You treat it like a teammate:
- Onboard it properly (define what it’s responsible for),
- Monitor performance (regularly review outputs),
- Adjust the process (tighten prompts, workflows, guardrails), and
- Audit outcomes (especially in high-stakes HR decisions)
This is especially critical as tools become more autonomous.
Defining AI in Three Buckets: Automation, Gen AI, and Agent AI
One part of the discussion that provided the most clarity about available tools was Trent’s framework for understanding “AI,” because not everything sold as AI is the same. But each tool can call into one of three buckets:
1. Automation: “Do this for me.”
Rule-based workflows and triggers (e.g., moving a candidate to a new hiring stage triggers an email). These are typically lower risk, but still need auditing.
2. Generative/conversational AI: “Work on this with me.”
Tools like ChatGPT, Claude, Copilot, or Perplexity that are collaborative and support writing, summarizing, brainstorming, and research. The human is still driving.
3. Agent AI: “Do this with me…and sometimes for me.”
This is where risk rises. Agents can take actions—not just generate text—and errors can become real-world outcomes.
Trent compared the best version of agent AI to an “Iron Man suit.” The human is still in charge, but dramatically augmented.
Where Should HR Start with AI?
Instead of beginning with tools, Trent recommended turning to journeys:
- The candidate journey
- The employee experience journey
- The hiring manager journey
Then, ask a simple set of questions at each step:
- Is this uniquely human? (Does it require trust, nuance, empathy, accountability?)
- If not, could AI or automation handle it?
- If yes, what governance is required?
This “decision tree” approach helps HR avoid both extremes:
- Over-automating meaningful human moments
- Ignoring obvious efficiency wins
A strong caution: don’t outsource the only human moments
Trent gave a pointed example around performance reviews. AI can absolutely assist in this area—but some vendors are pushing AI to write both the employee’s self-review and the manager’s review.
That should raise a red flag. For many leaders, performance conversations are one of the only true, focused 1:1 moments with employees all year. Outsourcing that entirely to AI can undermine trust and accountability.
Bias, Fairness, and Audits: A Pragmatic Take
A major audience question was: How do we ensure AI doesn’t discriminate in screening, assessments, or hiring decisions?
Trent offered a nuanced viewpoint:
- Bias exists today throughout human-centered processes.
- The question isn’t whether bias exists—it’s how well you can detect and audit it.
- In some cases, it may be easier to audit and correct bias in a system than in a human process.
But that only works if organizations put the right guardrails in place.
What governance should HR require from vendors?
Trent suggested asking vendors for evidence of responsible AI practices, including:
- a documented AI ethics framework (the “ten commandments” of what the tool will and won’t do),
- a cross-functional governance group to evaluate AI decisions and deployments, and
- third-party audits from experts who track regulations and best practices
The key takeaway: it’s hard to retrofit responsible AI after you’ve already scaled it.
Real-World AI Impact on HR: Two Examples HR Leaders Can Learn From
When asked where he’s personally seen practical impact, Trent shared two use cases that are especially relevant for HR:
1. Better leadership conversations through “personality-aware” prep
Using tools informed by leadership assessments (like DiSC or StrengthsFinder), he created a support system that helped tailor communication for difficult conversations.
The AI didn’t have human connection. Rather, it helped him prepare for it more effectively.
2. Anomaly detection in employee data
Instead of HR leaders manually combing through dashboards, an AI system flagged changes like, “Engineering in Costa Rica is down 25 points quarter over quarter.”
That kind of proactive signal allows HR to act faster—or at least pay attention sooner.
The Benefits of AI in HR: Driving Efficiency, Better Decisions, and Employee Experience
While much of the conversation around AI focuses on risks and hype, real value for HR teams lies in practical, measurable benefits. When implemented thoughtfully, AI can enhance human decision-making and significantly improve how HR operates.
1. Increased efficiency and time savings
AI helps automate repetitive, time-consuming tasks like resume screening, interview scheduling, and reporting. This allows HR professionals to focus on higher-value work, such as employee development, strategic planning, and culture-building.
2. More data-driven decision-making
AI enables HR teams to analyze large volumes of workforce data quickly and identify patterns that humans can miss. From predicting turnover risks to spotting engagement trends, AI supports more informed, proactive decision-making.
3. Improved candidate and employee experiences
AI-powered tools can streamline communication, personalize interactions, and reduce delays throughout the hiring and employee lifecycle, all of which help strengthen employer brand and engagement.
4. Enhanced talent matching and hiring outcomes
By analyzing skills, experience, and behavioral data, AI can help match candidates more effectively to roles. This can lead to better hiring decisions, improved retention, and stronger overall team performance.
5. Proactive issue detection
As highlighted in the discussion, AI can surface anomalies in workforce data, such as sudden drops in engagement or performance, allowing HR leaders to intervene earlier and address issues before they escalate.
6. Scalable learning and development
AI can personalize training recommendations and career development pathways based on individual employee needs, helping organizations plan and scale upskilling and reskilling efforts more effectively.
7. Stronger strategic impact for HR
By reducing administrative burden and improving insights, AI allows HR leaders to operate more strategically and speak the language of the business—whether that’s productivity, cost savings, or revenue impact.
Ultimately, the benefit of AI in HR isn’t about replacing people. It’s about augmenting their capabilities. Organizations that strike the right balance between automation and human judgment will be best positioned to drive both business performance and meaningful employee experiences.
Preparing Employees for AI: Reduce Fear by Increasing Control
Trent emphasized that employee anxiety about AI often isn’t just about jobs; it’s about loss of control.
When people don’t know what’s coming or what it means for them, the brain fills the gap with a story—and that story usually fuels fear.
His recommendation: bring employees into the process early.
- Let them experiment.
- Invite them into testing.
- Be transparent about what’s changing and why.
- Focus on reskilling and redeploying, not surprise displacement.
He argued that HR has to lead these conversations now—not after change has already happened.
“We’re Not Data-Ready”: The Simplest First Step (Even If It Isn’t Easy)
When asked how HR teams can use AI for workforce analytics when their data is scattered, Trent was blunt:
The first step is cleaning your data.
It’s not glamorous, but it’s unavoidable.
If dealing with a full organization’s worth of data feels overwhelming, start smaller:
- Pick one department,
- Clean and standardize that dataset,
- Build governance and SOPs, and
- Expand gradually once momentum and trust are established
Data Protection: What Should Never Go into AI?
One of the most practical moments came when Mira asked what people should avoid sharing, even if they think the environment is “safe.”
Trent’s answer:
If you wouldn’t want it read in court, don’t put it into a model.
If you must use AI for sensitive analysis:
- Remove or mask company identifiers (use “ABC Company”)
- Strip PII and confidential details
- Assume you may miss “one last detail” unless you’re extremely careful
And above all: HR shouldn’t do this alone. HR and IT/security need to be tightly aligned, with regular reviews of tools and policies.
ROI: How to Build a Defensible Business Case
To calculate ROI, Trent recommended using practical, finance-friendly math.
For example, if a recruiting tool saves 5 hours/week per recruiter, across 20 recruiters, that’s 100 hours/week regained.
Then translate the time saved into dollars and compare it to the tool’s cost.
He also recommended flipping the equation: if a tool improves speed or quality of hiring, estimate:
- reduced time-to-fill
- increased productivity
- revenue impact per filled role
His broader message was clear: HR leaders need to speak business language—not just HR language—to earn influence in AI decisions.
Choosing Vendors in a World Where Everyone Claims “AI”
Trent warned that some vendors market automation as “AI” to capitalize on hype. His advice:
- Know which type of AI you’re evaluating (automation vs. gen AI vs. agents)
- Start with one or two business problems—not 13
- Ask for proof: third-party audits, governance frameworks, transparency
Otherwise, it’s easy to get buried in endless demos and lose sight of what you’re actually trying to solve.
Recommended AI Tools to Explore (for Common HR Use Cases)
When asked about tools worth looking into, Trent mentioned:
- Perplexity for research and writing with sourced outputs (useful for credibility)
- Claude for complex writing and analysis tasks
- Google Labs tools such as Pomelli for brand-enabled content workflows (helpful for teams producing communications) and Gemini
He also shared a personal best practice: give your tool a knowledge base and rules, including pushing back on inconsistencies, so it acts like a rigorous editor—not just a friendly assistant.
Final Thought: AI Is a Leadership Challenge, Not a Tech Project
This conversation reinforced a theme many HR leaders are learning in real time:
Incorporating AI isn’t just a tool adoption cycle. It’s a people transformation, a process transformation, and a governance transformation.
The teams that are successful won’t be the ones who chase every shiny new product. They’ll be the ones who start with real problems, implement responsibly, and keep humans at the center of what matters most.
INTOO partners with HR leaders to bring forward-looking research, emerging trends, and practical best practices that drive real impact. If you’re preparing your workforce for AI adoption, we’re here to help—through manager training, resilience-building workshops, and programs that support internal mobility. Contact us to learn how we can be part of your transformation journey.
Robyn Kern is a seasoned business writer who has written in the HR, education, technology, and nonprofit spaces. She writes about topics including outplacement, layoffs, career development, internal mobility, candidate experience, succession planning, talent acquisition, and more, with the goal of surfacing workforce trends and educating the HR community on these key topics. Her work has been featured on hrforhr.org and trainingindustry.com.
