In my first article on artificial intelligence (AI) and workplace mental
health, I explored the promise and peril of AI-powered wellness tools:
chatbots, coaching apps, and digital screening platforms that aim to help workers manage
stress and identify mental health challenges. That conversation centered on a critical
question: Can AI help people cope better?
But there's a more fundamental question we should be asking: Can AI reduce the things that make coping so hard in the first place? This is where the conversation shifts from wellness—which implies helping workers adapt to difficult conditions—to work design, which means changing the conditions themselves.
The distinction matters enormously. A worker who uses an AI app to manage
anxiety while drowning in meetings, email, documentation, and conflicting priorities is
not experiencing burnout prevention. They are experiencing "productivity laundering,"
using a wellness intervention to mask the absence of systemic change.
AI has the potential to do something more fundamental: reduce the cognitive
overload, administrative drag, and unnecessary demands that wear people down in the
first place. Used thoughtfully, AI can become more than a productivity tool; it can be a
work-design tool that gives people back some of their mental bandwidth.
To understand what that looks like in practice, I spoke with Susan Frew,
Certified Safety Professional (www.susanfrewspeaks.com), an entrepreneur, AI
keynote speaker, and self-described AI "super user" who helps businesses integrate AI
into everyday workflows. Her message is refreshingly practical: The biggest gains do not
come from simply adding more technology. Instead, they come from first understanding how
the work gets done—where the friction lives, what can be automated, and where human
judgment still matters.
The Real Burden: What Actually Exhausts People
Burnout is not caused by everyday stress; burnout is caused by
chronic, unrelenting, and unnecessary stress combined with loss of control and
unclear purpose. Workers are depleted by the following.
Repetitive aspects. Administrative tasks that require cognitive effort but no expertise (documentation, reporting, data entry, scheduling, etc.).
Meeting bloat. Unnecessary meetings, badly run meetings, or meetings that could have been emails.
Inbox chaos. Not time management, but systems that make it impossible to prioritize.
Information hunting. Spending hours finding what should be easily accessible.
Decision fatigue. Too many routine decisions that don't require human judgment.
Role confusion. Unclear priorities, conflicting directives, or moving targets.
Context switching. Being interrupted so often that deep work becomes impossible.
These are not personal problem; they are design problems.
When organizations fix these problems, workers feel relief not
because they become more resilient, but because the burden itself shrinks. That is
fundamentally different from helping someone "cope" with exhaustion. Ms. Frew
explains it this way: You can't automate what you don't understand; the
organizations getting real relief from AI are the ones willing to do the work up
front—understand the workflow, identify what's creating friction, and decide where
human judgment still matters.
What AI Can Do: The Design Opportunity
AI excels at reducing cognitive load for exactly the kinds of tasks that are exhausting because they are repetitive, not because they are challenging. Consider the following.
Meeting summarization. A worker attends a
90-minute meeting; AI generates a summary, including decisions made, action
items, and who owns what. The cognitive load of "I need to remember this"
disappears. The worker spends minutes reading, not hours trying to reconstruct
what was said.
Drafting routine communications. A
customer service manager drafts the same type of email 20 times a week; AI
writes a first draft based on context, and the manager polishes it. Time saved:
4 hours per week. Cognitive load decrease? Huge!
Information arrangement. It takes a worker
45 minutes to search the company's systems for a policy, precedent, or past
decision; AI finds it in 90 seconds. Over a week: 4–5 hours reclaimed.
Workflow analysis. AI can identify
bottlenecks that humans experience as individual failures. "People are waiting
for approvals" isn't a person-management problem; it's a systems problem, and AI
can make it visible.
Scheduling and coordination. The
coordinator who spends 3 hours scheduling a meeting could use AI to handle the
logistics, and human judgment still decides whether the meeting is
necessary.
When AI is deployed against these specific burdens, something
important happens: Workers both work faster and experience actual relief; they have
brain space again.
What Burnout Prevention Could Actually Look Like
The mental health opportunity is this: When AI reduces unnecessary
cognitive load, workers can be redirected to work that engages the human capacities
that are most protective against burnout: autonomy, mastery, purpose, and
relationship. Imagine a workplace where the following occurs.
AI handles meeting summaries so managers can spend time on one-on-one conversations and coaching rather than administrative catch-up.
AI drafts routine documentation so people can spend time with customers
rather than record-keeping.
AI organizes information so people can spend time on analysis rather than hunting.
AI handles scheduling logistics so leaders can think strategically rather than managing calendars.
AI identifies workflow bottlenecks so teams can fix systems rather than blame individuals.
In each case, people are spending less time on depleting work and
more time on meaningful work. That is burnout prevention; Ms. Frew expounds further
below.
One of the biggest benefits I've experienced is getting the
open loops—all those unfinished tasks and "don't forget this" reminders
competing for mental space—out of my head. My AI system keeps track of projects,
people, commitments, finances, content, and follow-up so I don't have to keep
mentally rehearsing what I might be forgetting. That frees me to focus on the
work that needs me.
Productivity Laundering and the Speed Trap
This future depends entirely on leadership choices; the same AI
tools that can relieve burden can also be instruments of acceleration. Here is how
the danger plays out: A tax firm employs AI to assist attorneys in drafting routine
tax memos. Time savings per attorney: approximately 4 hours per week. However, the
firm does not reduce caseloads. Instead, the firm assigns each attorney 8–10
additional cases, and the 4-hour savings disappears. The attorney now works faster
but is no less exhausted. The technology gave a sense of efficiency but amplified
the underlying burden.
The pressure cooker effect.
Adding AI to already-stretched workplaces can often accelerate, rather than relieve,
burnout. The implicit message is that we have given you a tool to do more, so why
aren't you doing more? Workers feel it as pressure, not relief.
The new overload. AI tools
themselves create burden: There are too many platforms, too many logins, and too
many regular updates. Workers must become fluent immediately in a rapidly changing
tech ecosystem, and they may suffer "AI fatigue"—the mental toll of managing and
learning new tools—that is greater than the burden AI was meant to alleviate.
The Front-Loading Paradox
Employers need to anticipate another wrinkle: AI burnout
prevention has a J-curve. Before the technology starts giving people time and mental
bandwidth back, it may temporarily demand more of both.
Workflows must be mapped, and tools need to be selected and
configured. Information needs to be cleaned up, and employees need time to learn,
experiment, make mistakes, and figure out where AI genuinely helps. Systems need to
be integrated and adjusted, and only then does the cognitive dividend begin to show
up.
That means leaders should not abandon an AI initiative because it
has not reduced stress within a few weeks. But they also cannot simply pile the work
of AI adoption onto already-full jobs and call it innovation. Implementation is
work, and that work needs time, training, and breathing room. Ms. Frew explains
further below.
You can't just buy Copilot and expect everyone to be 25
percent more efficient next week. There is real front-loaded work in learning
AI. Leaders need to give people training, time to practice, and permission to
learn together—or the tool that was supposed to reduce stress simply becomes one
more thing employees are expected to figure out.
The irony is hard to miss: Organizations can create burnout on the
way to trying to prevent it. If AI is supposed to give
people time back, don't make them learn it on borrowed time.
The Digital Divide: Who Gets Left Behind?
AI may reduce the burden for workers who know how to use it, while
increasing pressure on those who do not. However, AI deployment may feel like a
threat, not support, for older workers, frontline workers, lower-wage workers, rural
workers, workers with disabilities, and those with limited tech access. They are
often unfairly labeled "resistant" or "not adaptable" when they never had the right
tools, training, time, or support.
Employers might consider how best to create the conditions for
everyone to adapt, which may include paid training time, clear pathways for help,
and psychological safety to ask questions without signaling that one is "not a
technology person."
Is this AI Burnout Prevention or Burnout Acceleration?
Organizations deploying AI should ask themselves the following questions.
What specific burden is this AI intended to reduce? (Not just "we want to increase productivity." That is a red flag.)
Will time saved be protected, reinvested in meaningful work, or swallowed by higher expectations? This is the decision point. Leadership must explicitly protect freed capacity or acknowledge that this is a productivity play, not a burnout prevention play.
Are workers being trained during paid work time? If workers must learn new systems on their own time, the burden has increased, not decreased.
Are managers ready to reset the norms around meetings, email, turnaround time, and availability? If AI takes out the meeting prep work but the meeting culture stays the same, the net effect is zero.
Are we measuring whether AI reduces burnout, or only whether it increases
output? These are fundamentally different. Output can increase while burnout
worsens, and burnout can decrease while output stays stable. Measure what you
care about.
Who might be excluded in AI implementation? And what are you actively doing to ensure no one is unfairly burdened by the transition?
Is this tool replacing bad management? (If so, it will fail. AI cannot fix toxic leadership, unclear direction, or excessive workloads.)
The Bottom Line: The Goal Is Not More Work Per Human
Here is the phrase that should guide every AI deployment decision
in the context of worker mental health: The goal is
not more work per human; the goal is more humane work.
More humane work means the following.
Less time on work that depletes without developing
More time on work that develops without exploiting
Clearer priorities and fewer false urgencies
Actual recovery time, not just scheduled time
Autonomy in how work gets done
Connection to purpose and impact
AI can support that vision when it serves as a work-design lever,
freeing human capacity to focus on meaning, skill-building, relationships, and
purpose. AI becomes a burnout accelerator when it serves only to increase pace,
pressure, ambiguity, and fear. The technology itself is neutral; the choice is
entirely human.
Opinions expressed in Expert Commentary articles are those of the author and are not necessarily held by the author's employer or IRMI. Expert Commentary articles and other IRMI Online content do not purport to provide legal, accounting, or other professional advice or opinion. If such advice is needed, consult with your attorney, accountant, or other qualified adviser.
In my first article on artificial intelligence (AI) and workplace mental health, I explored the promise and peril of AI-powered wellness tools: chatbots, coaching apps, and digital screening platforms that aim to help workers manage stress and identify mental health challenges. That conversation centered on a critical question: Can AI help people cope better?
But there's a more fundamental question we should be asking: Can AI reduce the things that make coping so hard in the first place? This is where the conversation shifts from wellness—which implies helping workers adapt to difficult conditions—to work design, which means changing the conditions themselves.
The distinction matters enormously. A worker who uses an AI app to manage anxiety while drowning in meetings, email, documentation, and conflicting priorities is not experiencing burnout prevention. They are experiencing "productivity laundering," using a wellness intervention to mask the absence of systemic change.
AI has the potential to do something more fundamental: reduce the cognitive overload, administrative drag, and unnecessary demands that wear people down in the first place. Used thoughtfully, AI can become more than a productivity tool; it can be a work-design tool that gives people back some of their mental bandwidth.
To understand what that looks like in practice, I spoke with Susan Frew, Certified Safety Professional (www.susanfrewspeaks.com), an entrepreneur, AI keynote speaker, and self-described AI "super user" who helps businesses integrate AI into everyday workflows. Her message is refreshingly practical: The biggest gains do not come from simply adding more technology. Instead, they come from first understanding how the work gets done—where the friction lives, what can be automated, and where human judgment still matters.
The Real Burden: What Actually Exhausts People
Burnout is not caused by everyday stress; burnout is caused by chronic, unrelenting, and unnecessary stress combined with loss of control and unclear purpose. Workers are depleted by the following.
These are not personal problem; they are design problems.
When organizations fix these problems, workers feel relief not because they become more resilient, but because the burden itself shrinks. That is fundamentally different from helping someone "cope" with exhaustion. Ms. Frew explains it this way: You can't automate what you don't understand; the organizations getting real relief from AI are the ones willing to do the work up front—understand the workflow, identify what's creating friction, and decide where human judgment still matters.
What AI Can Do: The Design Opportunity
AI excels at reducing cognitive load for exactly the kinds of tasks that are exhausting because they are repetitive, not because they are challenging. Consider the following.
When AI is deployed against these specific burdens, something important happens: Workers both work faster and experience actual relief; they have brain space again.
What Burnout Prevention Could Actually Look Like
The mental health opportunity is this: When AI reduces unnecessary cognitive load, workers can be redirected to work that engages the human capacities that are most protective against burnout: autonomy, mastery, purpose, and relationship. Imagine a workplace where the following occurs.
In each case, people are spending less time on depleting work and more time on meaningful work. That is burnout prevention; Ms. Frew expounds further below.
Productivity Laundering and the Speed Trap
This future depends entirely on leadership choices; the same AI tools that can relieve burden can also be instruments of acceleration. Here is how the danger plays out: A tax firm employs AI to assist attorneys in drafting routine tax memos. Time savings per attorney: approximately 4 hours per week. However, the firm does not reduce caseloads. Instead, the firm assigns each attorney 8–10 additional cases, and the 4-hour savings disappears. The attorney now works faster but is no less exhausted. The technology gave a sense of efficiency but amplified the underlying burden.
The pressure cooker effect. Adding AI to already-stretched workplaces can often accelerate, rather than relieve, burnout. The implicit message is that we have given you a tool to do more, so why aren't you doing more? Workers feel it as pressure, not relief.
The new overload. AI tools themselves create burden: There are too many platforms, too many logins, and too many regular updates. Workers must become fluent immediately in a rapidly changing tech ecosystem, and they may suffer "AI fatigue"—the mental toll of managing and learning new tools—that is greater than the burden AI was meant to alleviate.
The Front-Loading Paradox
Employers need to anticipate another wrinkle: AI burnout prevention has a J-curve. Before the technology starts giving people time and mental bandwidth back, it may temporarily demand more of both.
Workflows must be mapped, and tools need to be selected and configured. Information needs to be cleaned up, and employees need time to learn, experiment, make mistakes, and figure out where AI genuinely helps. Systems need to be integrated and adjusted, and only then does the cognitive dividend begin to show up.
That means leaders should not abandon an AI initiative because it has not reduced stress within a few weeks. But they also cannot simply pile the work of AI adoption onto already-full jobs and call it innovation. Implementation is work, and that work needs time, training, and breathing room. Ms. Frew explains further below.
The irony is hard to miss: Organizations can create burnout on the way to trying to prevent it. If AI is supposed to give people time back, don't make them learn it on borrowed time.
The Digital Divide: Who Gets Left Behind?
AI may reduce the burden for workers who know how to use it, while increasing pressure on those who do not. However, AI deployment may feel like a threat, not support, for older workers, frontline workers, lower-wage workers, rural workers, workers with disabilities, and those with limited tech access. They are often unfairly labeled "resistant" or "not adaptable" when they never had the right tools, training, time, or support.
Employers might consider how best to create the conditions for everyone to adapt, which may include paid training time, clear pathways for help, and psychological safety to ask questions without signaling that one is "not a technology person."
Is this AI Burnout Prevention or Burnout Acceleration?
Organizations deploying AI should ask themselves the following questions.
The Bottom Line: The Goal Is Not More Work Per Human
Here is the phrase that should guide every AI deployment decision in the context of worker mental health: The goal is not more work per human; the goal is more humane work.
More humane work means the following.
AI can support that vision when it serves as a work-design lever, freeing human capacity to focus on meaning, skill-building, relationships, and purpose. AI becomes a burnout accelerator when it serves only to increase pace, pressure, ambiguity, and fear. The technology itself is neutral; the choice is entirely human.
Opinions expressed in Expert Commentary articles are those of the author and are not necessarily held by the author's employer or IRMI. Expert Commentary articles and other IRMI Online content do not purport to provide legal, accounting, or other professional advice or opinion. If such advice is needed, consult with your attorney, accountant, or other qualified adviser.