AI in HR: How Human-Centered Leadership Can Close the Influence Gap at Work
Part of a series | Women@Work Series
Key takeaways
AI represents more than a technology shift. It represents a leadership, culture and work-design shift that requires intentional choices.
Human expertise matters more, not less. As AI generates more content and recommendations, leaders must help people strengthen judgment, critical thinking and contextual understanding.
Digital literacy must become continuous. Leaders don’t need to be experts in every tool, but they do need to build habits of safe experimentation and shared learning.
Burnout weakens performance and collaboration. When employees are overwhelmed, creativity, teamwork and decision-making suffer. Prevention and recovery are leadership responsibilities.
· Technology carries values. Every tool nudges behavior, so leaders should evaluate whether workplace technology supports the culture they want to build.
Systems shape influence. To broaden influence across the organization, leaders must look beyond individual confidence and examine norms, workflows, policies and decision-making structures.
The next leadership gap is an influence gap
At ADP’s 2026 Women@Work virtual summit, Virginia Magliulo, executive vice president, Employer Services International, ADP, offered HR and business leaders a timely challenge: Closing the leadership gap is less about who holds a title and more about who is valued, trusted, empowered and able to shape decisions inside the systems organizations create.
“Closing the leadership gap isn’t just about representation,” she said. “It’s about influence: who gets heard, trusted and empowered within the systems we have built.”
Work is changing quickly, but many organizations are still shaped by inherited norms about what leadership, expertise and authority should look like. At the same time, artificial intelligence (AI), digital collaboration tools and information overload are changing how people learn, decide and lead.
For HR and business leaders, there’s an opportunity to design work in a way that keeps people — and their expertise, well-being, judgment and relationships — at the center of transformation.
Why human-centered AI is now a leadership priority
Some organizations are adopting AI to move faster, increase efficiency and support decision-making. But as Digital Anthropologist and New York Times Bestselling Author Rahaf Harfoush emphasized during the session, the most important question is whether leaders are using it in ways that strengthen or weaken human capability.
AI can draft, summarize, analyze, recommend and automate, but if employees begin accepting AI-generated answers without questioning the assumptions behind them, organizations may gain speed while losing depth. If teams outsource too much thinking, they may become less practiced at the very skills leaders say they need most: judgment, creativity, problem-solving, collaboration and strategic thinking.
The risk is not that people use AI. The risk is that they use it passively.
According to ADP Research’s 2025 Global Workforce Survey, half of workers said they use AI at least multiple times a week, and one in five said they use it nearly every day. That level of adoption makes human-centered leadership more than a future-of-work ideal; it’s a current workforce priority.
Human-centered AI asks a different set of questions:
Does this tool help employees make better decisions?
Does it create space for human judgment?
Does it reduce friction without reducing accountability?
Does it support fairness, transparency and trust?
Does it help people build expertise or bypass it?
ADP Research also found that frequent AI users reported higher engagement, lower stress and more positive feelings about their teams, yet workers who use AI did not necessarily report feeling more productive. The finding underscores a key leadership challenge: AI adoption may change how work feels before it clearly changes how work performs.
For leaders, those questions should be part of every technology conversation, especially in HR. From hiring, recruiting and training to performance management, dashboards, chatbots and scheduling tools, AI-enabled technologies can shape how people experience work. That makes AI adoption a cultural decision rather than a systems decision alone.
“This is a leadership issue,” Harfoush said. “This is not a tech issue. This is not an AI issue. It’s not about the tools, but how we’re using these tools.”
Launch the virtual session anytime
Closing the leadership gap: Unlocking women’s full influence in 2026
Protect human expertise in an age of instant answers
One of Harfoush’s central points was that the workplace is shifting from a search culture to a generation culture. In the past, employees searched across sources, compared information and formed conclusions. Increasingly, people ask a tool and receive a polished answer.
That shift can be helpful, but it can also change how people learn. The process of searching, comparing, questioning and revising can be an educational experience that helps people build discernment.
The speed of AI-generated work can also create a false sense of confidence.
“How do we actually know if it’s good?” Harfoush said. “Is the data good? Is the timeline realistic? Is the demographic analysis accurate?”
These questions are especially important when employees use AI tools. A strong AI output can look authoritative, even when it needs review. Leaders should normalize the idea that AI-generated work is a starting point, not an endpoint.
HR and business leaders can help protect discernment by creating a culture where employees are expected to explain their thinking, rather than just present their outputs.
That same principle applies to AI-supported hiring.
Instead of asking only, “What did you produce?”, leaders can ask:
What assumptions shaped this recommendation?
What did you challenge or verify?
Where could this answer be incomplete?
What human context matters here?
What would change your mind?
“AI is supercharging deep expertise,” Harfoush said, “but it’s collapsing weak knowledge systems.”
The organizations that benefit most from AI will likely be those that pair it with deep human expertise.
ADP Research’s People at Work 2025 research highlights the urgency of that investment: Only 24% of the global workforce said they were confident they had the skills needed to advance to the next job level in the near future, and just 17% strongly agreed their employers were investing in the skills they need for career advancement.
That means investing in people’s ability to evaluate quality, recognize nuance, spot bias and understand the consequences of decisions.
ADP Research has also found that less than 4% of workers were upskilled in their first two years of employment, based on an analysis of payrolls of 51 million private-sector employees between 2019 and 2023. For leaders, that gap points to a practical opportunity: Build expertise into the flow of work before AI changes the work faster than people can adapt.
Build digital literacy through safe experimentation
Digital literacy may have been a training module in the past, but now, it’s an ongoing leadership practice.
No leader can keep up with every tool, platform or feature; however, leaders can create an environment where teams learn how to experiment thoughtfully.
“The skill that I’m talking about is a mindset, and that is the mindset of embracing uncertainty through active experimentation,” Harfoush said.
Experimentation shouldn’t equate to random tool usage, however. It should equate to structured learning.
A practical workplace experiment might include a specific process to test, an approved tool, a safe environment, no sensitive information, a defined time period and a clear plan to assess whether the tool helps.
For example, a team might test whether an approved AI tool can help draft first-pass learning materials, summarize internal knowledge or compare communication approaches. The goal is not to declare the tool “good” or “bad.” The goal is to understand how it behaves, where it helps, where it falls short and what human oversight it needs.
HR can play a foundational role here by helping teams create safe experimentation norms. That includes clarifying which tools are approved, what data can be used, when human review is required and how employees should disclose AI-assisted work.
“A culture of experimentation gives you the confidence to take a specific amount of observable risk so that you’re not putting your data, your team, your technology in danger,” Harfoush said.
The cultural benefit is just as important. When experimentation is done well, it reduces fear, encourages curiosity and promotes shared learning.
Learn how to create positive employee experiences for better business outcomes.
Reevaluate speed as the default measure of success
AI often enters organizations with a promise of speed: faster drafting, faster analysis, faster service, faster decisions.
Speed can be valuable, but speed without judgment can create rework, risk and mistrust.
ADP Research found a similar tension in workers’ experience with AI: Daily AI users were four times as likely as nonusers to say they were less productive than they could be, even though frequent AI users also reported higher engagement, lower stress and more positive feelings about their teams.
Harfoush challenged leaders to rethink what they are truly optimizing for.
“Instead of optimizing for speed, we’re going to optimize for collaboration,” she said. “We’re going to optimize for depth. We’re going to optimize for connection. We’re going to optimize for creativity.”
That challenge is highly relevant for HR and business leaders designing modern work. If the organization rewards only speed and volume, employees may feel pressured to produce more without enough time to think, collaborate or verify.
That can be especially important in high-impact people processes such as hiring, recruiting, training, performance management and scheduling, all of which require context. AI can support those processes, but leaders should be careful not to confuse efficiency with effectiveness.
A better question is: Where should we move faster, and where should we slow down to protect quality?
Leaders can make that distinction visible by asking which tasks are appropriate for automation, which require human review and which need deeper discussion because they involve judgment, context, relationships or risk.
This framework helps teams use AI with intention rather than applying speed everywhere.
Address burnout as a performance and culture issue
The future of work is often discussed in terms of technology, but Harfoush reminded leaders that humans matter. Employees are people navigating pressure, uncertainty, screens, notifications and emotional demands throughout the workday.
Burnout is more than fatigue or exhaustion. It can affect how people collaborate, interpret tone, take risks, learn, innovate and engage.
ADP Research’s global workforce data shows how much room there is to improve engagement: Fewer than one in five workers worldwide were fully engaged on the job in 2025.
A burned-out workforce is less likely to have the energy required for thoughtful transformation.
“Think of burnout as a thief,” Harfoush said, “a thief that steals energy and motivation, a thief that steals empathy and a thief that steals confidence.”
That makes recovery a leadership responsibility.
ADP’s 2026 HR trends research found that a majority of organizations (74% or more) agreed they have a responsibility to ensure employees’ physical, mental and financial well-being, while also noting a gap between that sense of responsibility and confidence in their ability to offer the resources needed to enhance well-being.
HR and business leaders can start by reviewing how work is designed:
Are calendars filled with back-to-back meetings?
Do employees have time to think before making decisions?
Are urgent requests truly urgent?
Are people expected to respond outside working hours?
Are managers modeling healthy boundaries?
Are technology tools reducing work or simply adding more channels?
Teams can limit the number of meetings per day, avoid back-to-back meetings, build transition time into calendars, protect time to think and plan recovery after intense work cycles. Leaders can also model healthier communication norms by being explicit about response expectations.
The goal is not to eliminate pressure, as some pressure is part of growth. The goal is to prevent chronic overload from becoming the prevailing structure of the organization.
“If you are in fight or flight, if your team is in fight or flight, they cannot be creative, be collaborative, be strategic, be agile,” Harfoush said.
Audit the beliefs embedded in workplace technology
One of the most interesting ideas from Harfoush’s discussion is that technology reflects belief systems.
“All technology is the manifestation of belief systems,” she said.
Every platform, dashboard and tool is designed around assumptions about what matters.
A scheduling tool may assume productivity means optimizing every open minute. A dashboard may assume output is best measured by activity. A document tool may assume summaries are always better than deep reading. A performance dashboard may emphasize hours worked while missing context such as emotional labor, complexity or the different kinds of energy different tasks require.
These design choices influence workplace culture.
“When you bring a tool into your organization, you’re not just getting the technical features and the benefits,” Harfoush said. “You’re also getting that belief system.”
That’s why HR leaders should not evaluate technology solely by its features, but by the behaviors each tool encourages.
ADP’s 2026 HR trends report identifies responsible AI governance, transparency and HR-IT collaboration as increasingly important to shaping the future of work. The report also notes that as AI transforms the workplace, leaders are using data and technology to identify key competencies and strategically redesign roles to align talent with business needs.
Before implementing a new workplace technology, leaders can ask:
What does this tool define as success?
What behaviors does it reward?
What does it make visible or invisible?
Who might benefit from this design?
Who might be disadvantaged by it?
Does it support our stated values?
These questions are especially important as AI becomes embedded in more HR processes. If a tool shapes hiring, development, scheduling, coaching or performance decisions, it must be assessed not only for capability but also for alignment with organizational culture.
Expand influence by redesigning the system
The session’s broader theme centered on unlocking influence at work. Harfoush’s message suggests that influence is created when systems, norms and leadership practices enable more people to contribute meaningfully. That requires leaders to examine how decisions get made.
Consider questions like:
Who is invited into early-stage strategy conversations?
Whose expertise is recognized?
Which behaviors are rewarded in meetings?
Are quieter or more reflective employees given ways to contribute?
Are decisions made only in live meetings, or can people contribute asynchronously?
Are managers trained to identify and reduce bias in AI-supported processes?
Are employees encouraged to challenge outputs, assumptions and defaults?
Influence grows when leaders create structures that make contribution easier, safer and more visible. That might include creating groups such as innovation squads, hosting mental gym sessions, encouraging safe experimentation, building recovery time into calendars and auditing whether workplace technologies align with organizational beliefs.
For women leaders in particular, Harfoush encouraged confidence in leadership styles that may not match outdated archetypes.
“The way that I lead might not look like the traditional archetype of leadership, but it’s still just as effective,” she said.
For HR, this is a chance to help organizations move from representation alone to participation, voice and impact.
A human-centered leadership agenda for 2026
The pace of change might not slow down, but leaders can choose how their organizations respond.
A human-centered leadership agenda for 2026 should include five priorities:
Invest in expertise. Help employees build the judgment needed to use AI well.
Create safe experimentation. Encourage curiosity within clear guardrails.
Redesign work for focus and recovery. Treat attention and energy as strategic resources.
Evaluate technology for cultural fit. Go beyond features to the behaviors that tools encourage.
Broaden influence intentionally. Design systems that enable more people to contribute, challenge and lead.
Many employers appear to be approaching AI as an augmentation strategy rather than a replacement strategy. According to ADP’s 2025 HR trends survey, 84% of large organizations, 76% of midsized organizations and 73% of small organizations agreed that AI can help streamline processes but will not replace employees.
As Harfoush said, culture is built by choice, not by chance. AI will continue to reshape work, but leaders still shape the conditions in which work happens. The organizations that thrive will be those that do not simply adopt new tools but build human capacity to use them wisely.
And when the pace of change makes the path forward unclear, leaders can return to one of Harfoush’s closing reminders.
“When we know the answers are always changing, the only thing we can do is get really, really good at asking the right questions.”
Learn more
Explore ADP resources to help your organization build a more human-centered, resilient and future-ready workforce.
FAQs
What is human-centered AI in the workplace?
Human-centered AI in the workplace is the practice of using AI in ways that support people’s judgment, well-being, fairness and effectiveness. It keeps humans accountable for decisions and treats AI as a tool for augmentation rather than a replacement for expertise.
How can HR leaders use AI responsibly?
HR leaders can use AI responsibly by setting clear governance standards, protecting employee and candidate data, requiring human oversight, monitoring for bias and training employees to evaluate AI outputs critically.
Why is digital literacy important for leaders?
Digital literacy helps leaders understand how technology shapes decisions, behavior and culture. Leaders do not need to become technical experts, but they do need enough fluency to ask better questions, assess risks and guide teams through change.
How can organizations reduce burnout while adopting new technology?
Organizations can reduce burnout by simplifying workflows, limiting unnecessary meetings, clarifying communication expectations, creating focus time and ensuring technology simplifies work rather than adding complexity.
What leadership skills matter most in the future of work?
The future of work requires critical thinking, adaptability, empathy, communication, ethical judgment, collaboration and the ability to lead through ambiguity. These skills help organizations balance innovation with trust.
How does AI affect workplace culture?
AI affects workplace culture by changing how people make decisions, communicate, measure productivity and define expertise. If implemented without intention, it can reinforce unhealthy norms. If implemented thoughtfully, it can support better learning, connection and performance.
