How to Protect Human Creativity As AI and Automation Scale

With AI and automation becoming part of everyday work, the leadership challenge is no longer simply deciding what can be automated. It’s deciding what should be. As these systems become more capable, they can take on work that once required significant human effort—but greater efficiency does not necessarily lead to better thinking, stronger decisions or more innovative outcomes.

The organizations that gain the most from AI may be those that use it not simply to increase output but to create more capacity for distinctly human work. Gallup’s 2025 research on AI use at work found that the share of U.S. employees using AI at work at least a few times a year nearly doubled in two years, rising from 21% in 2023 to 40% in early 2025. As adoption accelerates, leaders must consider how to use technology in ways that strengthen rather than diminish human capabilities.

That question is at the center of insights from members of the Senior Executive HR Think Tank. Drawing on experience in leadership development, HR, organizational psychology, human performance and business transformation, they argue that the strongest human-AI models are not built around a fixed division between people and technology. Instead, they capitalize on the different strengths each brings to the work—using automation to create capacity and deliberately reinvesting it in judgment, creativity, relationships and strategic thinking.

“Automation handles the distance. You handle the direction. What I see consistently: Leaders who stay curious use AI to think bigger. Those who stay busy use it to move faster.”

– Aurelien Mangano, Trusted Leadership Advisor at DevelUpLeaders

SHARE IT

Let Automation Create Capacity for Better Thinking

For Aurelien Mangano, Trusted Leadership Advisor at DevelUpLeaders, technology itself is not the starting point. The quality of leadership is.

Mangano says leaders tend to fall into two camps when using AI: those who use it to expand their thinking and those who use it to accelerate their existing workload.

“Automation handles the distance. You handle the direction,” Mangano says. “What I see consistently: Leaders who stay curious use AI to think bigger. Those who stay busy use it to move faster.”

That distinction puts creativity in a different category from routine productivity. If automation allows organizations to produce more of the same work faster, its value remains limited. Leaders can create more space for creativity by using technology to handle execution while retaining responsibility for direction, context and purpose.

“The difference is always the quality of thinking behind it,” Mangano says.

The implication for executives is straightforward: Automation should not become an excuse to eliminate the thinking that gives work its strategic value. It should create room for leaders and employees to do more of it.

Decide What Must Remain Human

Michelle Arieta, Chief People Officer and Consultant at Polaris Pathways, takes the argument further: Leaders should stop thinking about automation and creativity as forces that need to be kept in equilibrium.

“I don’t think ‘balance’ is the right word,” Arieta says. “Balance implies two forces in equilibrium you can dial against each other. Automation capability compounds; human bandwidth for judgment doesn’t.”

Her alternative is to make an explicit distinction between decisions that should remain human and work that can be automated aggressively.

Arieta points to consequential moments such as compensation decisions involving struggling high performers, decisions about employees in their first 90 days and reorganizations that may affect people’s relationships and livelihoods.

“Those need a human who’s accountable and can sit with getting it wrong,” she says.

This distinction aligns with a broader concern about how organizations redesign decision-making around AI. Deloitte’s 2026 Global Human Capital Trends research argues that technology can accelerate analysis and help clarify uncertainty, but human purpose, values and judgment remain essential to consequential decisions. The research warns that organizations that fail to deliberately design for human agency risk diluted accountability and the erosion of human involvement in decisions where it still matters most.

Routine activities—including scheduling, screening and data pulls—can then be automated without treating automation as a threat to creativity. In Arieta’s framework, the opposite is true: Removing repetitive work creates the capacity people need to think, create and make decisions.

“That’s not a tradeoff against creativity, it’s what makes room for it,” she says.

That requires leaders to establish clear boundaries around consequential decisions rather than assuming every task will eventually become appropriate for automation.

“Automation excels at execution—repeatable, rule-based work. Creativity excels at framing, deciding what’s worth solving and when to break the rules.”

– Maureen Metcalf, Founder and CEO of the Innovative Leadership Institute

SHARE IT

Give Humans More Space to Think

Maureen Metcalf, Founder and CEO of the Innovative Leadership Institute, frames automation and creativity as a governance challenge.

“Automation and creativity aren’t a one-time tradeoff—the balance shifts as automation capability grows, so it needs ongoing governance,” she says.

One of her key distinctions is between execution and framing. Automation can be highly effective at repeatable, rule-based work, while humans remain responsible for deciding what is worth solving and when to challenge the established rules.

“Automation excels at execution—repeatable, rule-based work. Creativity excels at framing, deciding what’s worth solving and when to break the rules,” Metcalf says.

That distinction also changes how organizations should measure productivity. If automation allows employees to process more work, leaders may see efficiency gains without realizing any creative benefit. Protect freed capacity for higher-value thinking.

Metcalf recommends going “tight” where variability adds little value, such as compliance and reporting, and “loose” where variation creates value, such as strategy and ideation. She also recommends using automation to generate first drafts and options while reserving selection and judgment for humans.

“Protect unstructured time deliberately—it doesn’t survive by default at scale,” she says.

A broader trend identified in the 2025 Stanford AI Index Report supports Metcalf’s point. The report found that organizations were rapidly expanding their use of AI, with 78% of surveyed organizations reporting AI use in 2024, up from 55% the year before. The report also found significant productivity gains across a range of tasks. But greater productivity does not automatically determine how organizations should use the capacity those gains create.

That is where leadership decisions matter. Organizations can absorb newly available capacity by demanding more volume, or they can reinvest some of it in strategic thinking, experimentation and problem-solving.

“Being stuck in an automated process with no way to reach a human is one of the biggest sources of customer and employee frustration.”

– Tracy Jackson, President and CEO of HR E-Z

SHARE IT

Design the Human Escape Hatch

Tracy Jackson, President and CEO of HR E-Z, approaches the question from human-centered design. Jackson says the first question should not be what technology can automate but what people actually need from the process.

“Balance starts with human-centered design: automation built starting from human end-users’ needs and ending with human outcomes, not the other way around,” She says.

That means automating repeatable and trackable work like data entry, routing and information compilation while preserving human judgment and relationships around how people interact with the resulting information.

Jackson also emphasizes data quality. “Good info in, good info out,” she says, warning that automated systems need to be tested against their original intent rather than evaluated solely on efficiency.

Perhaps most importantly, she says organizations need an “escape hatch”: a clear way for employees, customers or other users to reach a person when automation fails to address their needs.

“Being stuck in an automated process with no way to reach a human is one of the biggest sources of customer and employee frustration,” Jackson says.

This focus on human agency also reflects Deloitte’s 2026 research, which highlights the importance of designing human involvement into AI-supported decision-making rather than treating people as passive recipients of automated recommendations.

Human involvement, then, should not be reserved only for emergencies. It should be deliberately built into workflows where context, empathy, interpretation and judgment affect the outcome.

Build Intentional Pauses Into AI Workflows

Dr. Ariel Shivers McGrew, Business Psychologist and Founder of Tactful Disruption®, approaches automation through the lens of cognitive load and behavioral health. McGrew says the behavioral health field offers an important lesson: Automation can improve performance not merely by speeding work up but by deliberately creating moments when people stop and think.

“The smartest systems now build in stop and go checkpoints that catch rushed entries before they become quality issues,” McGrew says.

Those checkpoints can be small. AI workflows can require brief edits, text-box reflections or confirmations before a task is submitted. The purpose is not to slow work unnecessarily but to create a moment for human judgment.

“When AI workflows require brief edits, text-box reflections, or quick confirmations before submission, you don’t just improve documentation—you improve thinking,” she says. The approach also creates a feedback loop. Better human input produces cleaner data, which can make automated systems easier to scale because they spend less time correcting poor inputs.

McGrew’s recommendation challenges the assumption that every additional human touchpoint represents inefficiency. In some processes, a deliberate pause may protect quality, strengthen judgment and create an opportunity to challenge weak output before it becomes a larger problem.

“By designing AI to slow people down just enough to be intentional and speed them up everywhere else, you learn 10 times more, strengthen quality control and make creativity possible again,” McGrew says.

For leaders, that means automation design should include deliberate moments of reflection rather than treating human involvement as friction that must always be removed.

Redesign Work Around Human Strengths

Amy Douglas, Chief, Culture and Connection at Levata Human Performance, takes a systems-level view. Douglas says the central issue is not whether organizations should automate or preserve creativity. It is whether they are redesigning work so humans and AI contribute where each performs best.

“At Levata, our Human + AI Performance System starts with the belief that AI amplifies the human system it lands in,” Douglas says. “When people bring creativity, strong judgment, critical thinking and connection, AI accelerates performance.”

The reverse is also true. If those capabilities are weak, technology can accelerate poor decisions and reactive behavior.

“The organizations seeing the greatest value from AI aren’t trying to automate human creativity,” Douglas says. “They’re redesigning work so AI handles repetitive and analytical tasks while humans focus on problem framing, judgment, innovation, collaboration and meaning-making.”

Her point of view is consistent with the broader picture emerging from Stanford’s 2025 AI Index: AI is becoming more deeply embedded in business, and evidence of productivity gains is growing across different types of work. The strategic question for leaders is increasingly less about whether AI will enter the workflow and more about how work should change once it does.

For HR leaders, that can mean identifying which capabilities become more important as automation expands and then deliberately developing them. Creativity, judgment, connection and critical thinking cannot simply be assumed to survive an automated workplace. They need to become part of how work is designed, managed and evaluated.

“At scale, the goal isn’t more automation or more creativity,” Douglas says. “It’s creating capacity through automation and reinvesting that capacity into higher-value human work.”

Turn Automation Into A Creativity Dividend

The members’ advice points to a common principle: Automation creates the greatest value when organizations decide in advance what they intend to do with the capacity it creates.

• Use AI to expand thinking, not simply accelerate output. The value of automation depends on the quality of human thinking directing it.
• Define which decisions must remain human. Establish clear boundaries around decisions that require accountability, context, empathy or judgment.
• Protect the time automation creates. If freed capacity is immediately consumed by additional volume, organizations gain efficiency without gaining creativity.
• Design human escape hatches into automated processes. Give employees and customers a clear path to human support when automated systems cannot resolve their needs.
• Build intentional pauses into AI workflows. Brief reflection points can improve judgment, data quality and learning without eliminating the efficiency benefits of automation.
• Redesign work around complementary strengths. Let AI handle repetitive and analytical tasks while humans concentrate on framing problems, making judgments, innovating and building relationships.

The Competitive Advantage Is Human-Centered Work Design

The most effective approach to automation is not a fixed formula for dividing work between humans and machines. It is an operating model that evolves as technology changes. Leaders must continually reassess which work benefits from consistency and which requires judgment, creativity or human connection.

The organizations that gain the most from AI will not necessarily be those that automate the most. They will be those that look beyond efficiency gains to understand what people can do when technology is handling something else—and know how to turn that capacity into higher-value human work.