AI Adds Capacity. People Don’t Scale the Same Way: Lessons on Sustainable Operations

Author: Ellenit Serrano

AI has become the stated reason for a new wave of restructuring. In May 2026, AI was the most-cited reason for US job cuts, behind about 40% of them. Most of the conversation centers on which roles will change, and how fast. Far less attention goes to a simpler question: what happens to the people inside the change while it is underway?

For every leader, that gap is where the work begins. Every operating model creates a human experience, whether intentional or not, and a restructuring or an AI rollout is an operating model change. This article draws on recent research and my experience as an executive leader to offer six practical moves for protecting people and results through change.

The Load No Plan Accounted For

I saw this gap firsthand earlier in my corporate career. During an organizational and geographic restructuring that ended with the closure of an office home to a few hundred people, layoffs came in waves. Every day, people from across the office walked into mine. They weren’t there for a meeting. They came to talk; somewhere it felt safe to say what they couldn’t say in a team meeting. Nothing in the transformation plan accounted for it.

The plan also couldn’t see where I was starting from. A young mother in a demanding senior role, I was already on the verge of burnout, living through the same uncertainty as everyone else, while absorbing theirs. The restructuring didn’t create my burnout; it took me past the edge. The people others lean on are often the ones already carrying the most, and transformation plans rarely account for them.

The Restructuring That Doesn’t End

That restructuring I experienced earlier in my career had a beginning and an end. AI-driven change often doesn’t. It arrives in waves, reaches every function, and asks people to adopt the very tools that put their roles in question.

In INTOO’s Employee Mindset Report, conducted by The Harris Poll, the share of employed Americans concerned about being laid off rose from 48% in 2019 to 61% in 2025. Nearly half (49%) said they would trust employees’ social media posts over CEO announcements for the real story behind layoff news. That is the modern version of my office door: when official channels go quiet, people find somewhere else to process what is happening. It also signals psychological safety, the shared belief that it is safe to ask questions or admit a struggle. When jobs feel at risk, that belief can erode, and the real conversations move out of the room.

Mercer’s Global Talent Trends 2026 found that the share of employees who say they are thriving fell from 66% in 2024 to 44% in 2026. In Korn Ferry’s Workforce 2026 Global Insights Report, 45% say they are too busy to deliver meaningful results that contribute to growth. That is the bottom-line cost: a transformation launched to create value, carried by people who no longer have the capacity to deliver it.

Who Carries the Load

The people who remain are running the business and changing it at the same time. In the same Korn Ferry survey, 52% say AI tools have increased the number of tasks expected in their role. Among remaining managers, the people others lean on most, 55% say they are exhausted.

Research by Boston Consulting Group and the University of California – Riverside, published in Harvard Business Review, helps explain why. It links “AI brain fry,” mental fatigue from excessive use or oversight of AI tools, to more errors and higher intent to quit. Inside that strain sit the informal anchors: the managers and colleagues others walk in to talk to. A restructuring adds to their load first and notices it last. AI also raises the stakes on speaking up. Saying a new tool has added work can feel like making the case for one’s own replacement, so strain can stay hidden until it shows up as errors, disengagement, or resignations.

Safety Starts With a Steady Leader

Direct leaders shape the daily experience of work more closely than any program can. In a survey by The Workforce Institute at UKG, 69% of employees said their manager affects their mental health, the same share who said so of a spouse or partner. A 2023 meta-analysis of 53 studies linked transformational, relationship-focused, and task-focused leadership to better employee mental health, and destructive leadership to worse.

Psychological safety is felt before it is stated. Research on emotional contagion shows that moods transfer among group members and shape cooperation and conflict. During a restructuring, leaders are under strain too, and that strain travels: the tight voice, the rushed meeting, the answer that comes too fast. A leader who is visibly on edge signals threat, and people stop speaking up.

Simple practices can help a leader steady before a hard conversation. A 2023 study found that breathwork may lower stress, anxiety, and depression. From that steadier place, a leader can say “I don’t know yet” without passing on anxiety, listen without rushing to reassure, and stay present while the answers are still forming. None of this replaces good decisions or clear communication. It makes them more possible under pressure.

Capacity Gained, Capacity Spent

Organizations are adopting AI to expand capacity, and that gain is real. But it rests on something that doesn’t scale the way software does: the attention, energy, and judgment people bring to their work are finite. The brain fry study linked using AI to offload routine work with lower burnout, while constantly overseeing AI tools raised mental strain. So, the productivity case and the people case aren’t in conflict. Productivity gains that burn through human capacity don’t last. They resurface later as errors, disengagement, and the loss of the people an organization most needs. The question isn’t whether to use AI to grow capacity. It is who decides where the gain goes.

Six Moves Toward Sustainable Operations

Sustainable operations put human sustainability into practice: operating models designed to keep delivering results while strengthening the people who make those results possible. Protecting people through change isn’t a wellness program running beside the transformation. It is part of how the transformation is run.

1. Put human capacity in the business case. Most organizations already have the data: engagement surveys, overtime, unused leave, attrition. Bring it to the decision. Before approving a restructuring or an AI rollout, ask the same question asked of budget and systems: is there capacity to absorb this, and where is the organization already running hot?
2. Decide where the capacity gain goes. When AI frees time, decide explicitly how it is spent: some on more output, some on retraining people for the work AI is changing. Retraining needs protected time; training added to an already full week becomes one more task. Involve the people doing the work in redesigning it. For every new tool, integration step, or reporting requirement, name what stops or waits. Without a decision, the default is more volume.
3. Shorten uncertainty. Sequence people decisions early and share timelines, even provisional ones. Say plainly what is not yet known. Silence fills with rumor.
4. Make it safe to say “this is too much.” Leaders go first: they name their own limits and ask openly what should stop. Thank the people who raise workload or AI concerns, and act on what they say. If people only speak up in private, the organization learns too late.
5. Equip managers and support the people others lean on. Identify informal anchors early. Give them information first, time to listen and support of their own. Give them clear referral pathways so they are not the only support. Check in on them, not only through them. Senior leaders need the same support.
6. Measure strain next to milestones. Put a few human indicators on the same dashboard as adoption rates and savings targets: regretted attrition, overtime, sick leave, and a short pulse asking whether people feel safe raising concerns. Name who owns them, and build them into leaders’ goals.

Conclusion

AI can take on more of the routine work. What it can’t supply is human capability: the judgment, creativity, relationships and trust that turn a tool into results. And capability depends on capacity. People who are depleted, anxious, or afraid to speak up can’t bring their best judgment to the work. Start with one question at the next steering committee review of an AI rollout or restructuring: who is carrying this, and what is coming off their plate?

About the Author

Ellenit Serrano is a workplace wellbeing strategist and enterprise operator who helps leaders build sustainable operations, where people and performance reinforce each other. She is Founder and Advisor at Esencia Solutions and Vice Chair of the Global Wellness Institute’s Workplace Wellbeing Initiative. She previously held senior leadership roles at American Express and Mastercard.

*** Disclaimer ***

The blog submissions featured on this site represent the research and opinions of the individual authors. The Global Wellness Institute and the Workplace Wellbeing Initiative are not responsible for the content provided. The views expressed are solely those of the authors and do not necessarily reflect the official policy or position of the Global Wellness Institute or the Workplace Wellbeing Initiative. Readers are encouraged to consult with a qualified healthcare professional for specific health concerns.