Business leaders can buy powerful artificial intelligence tools, provide employees access, and still fail to generate the results they expected.
Why?
The problem may not be the technology.
It may be the psychology of the people being asked to use it.
That was one of the central themes of my recent conversation with Dr. Gleb Tsipursky, behavioral scientist, CEO of Disaster Avoidance Experts, author of seven books, and author of The Psychology of AI Adoption at Work: From Resistance to Results.
Dr. Tsipursky has spent more than 15 years in academia, including faculty appointments at The Ohio State University and UNC-Chapel Hill. His work and insights have appeared in Harvard Business Review, Fortune, Forbes, and The New York Times.
His message to CEOs, business owners, executives, HR leaders, and managers is important:
Successful AI adoption is not simply a technology implementation problem. It is a leadership, behavior, and psychology problem.
Why AI Adoption Is Different From Previous Technology Changes
Business leaders have managed technology transitions before.
Companies moved from paper records to computers. Teams replaced spreadsheets with customer relationship management systems. Email changed the way organizations communicated. Enterprise software transformed accounting, HR, operations, and sales.
Those changes sometimes created resistance because employees had to learn unfamiliar systems.
AI introduces another layer of concern.
An employee learning a new CRM generally does not believe the CRM is being trained to replace them.
Artificial intelligence can feel very different.
Employees may look at an AI system that can write emails, analyze data, generate reports, create images, research prospects, or perform pieces of their workflow and ask a much more personal question:
“Am I being asked to train my replacement?”
That changes the psychology of adoption.
The 3 Psychological Barriers to AI Adoption
Dr. Tsipursky identified three important emotional barriers leaders should understand.
1. Fear and Anxiety About Job Loss
Employees may believe greater AI efficiency will eventually lead to fewer jobs.
Telling them not to worry does not automatically eliminate that concern.
Executives who are enthusiastic about artificial intelligence may see increased productivity, faster workflows, and opportunities for growth.
An employee may see the same technology and wonder whether their position will exist two years from now.
Ignoring that fear can make resistance worse.
Leaders need to address it directly and explain how AI fits into the company’s workforce strategy.
Where circumstances allow, Dr. Tsipursky recommends positioning AI around growth rather than simply cost cutting.
The message becomes less about replacing people and more about increasing what capable employees can accomplish.
2. Threats to Professional Identity
Job security is not the only issue.
AI can also threaten the work employees take pride in.
Consider a salesperson who considers prospecting a core professional skill.
A writer may value the ability to communicate persuasively.
An analyst may take pride in building financial models.
A marketer may identify strongly with creating effective campaigns.
When leadership introduces AI specifically to perform those tasks, the employee may experience the technology as a threat to their professional identity.
The reaction is not necessarily, “This software is difficult to use.”
It may be:
“If AI does the part of my job I am proud of, what is my value?”
That distinction matters.
Why Leaders Should Start With the Tasks Employees Hate
One of the most practical ideas from our conversation was surprisingly simple.
Do not begin an AI initiative by asking:
“What valuable employee work can we automate?”
Start by asking employees:
“What are the three tasks you hate doing?”
Look for repetitive, annoying, time-consuming activities employees would gladly remove from their workload.
Examples might include:
- Gathering information from multiple systems
- Preparing recurring reports
- Summarizing documents
- Searching for information
- Preparing meeting materials
- Organizing notes
- Drafting routine communications
- Reformatting information
- Performing repetitive administrative work
This approach can change the employee’s first experience with AI.
Instead of AI taking away something they value, AI removes something they dislike.
That creates an opportunity for employees to experience the technology as useful before leaders ask them to apply it to more important parts of their work.
The AI Sales Email That Hurt Customer Trust
Dr. Tsipursky shared an example involving an investment management company using AI for prospect outreach.
The system could identify potential prospects based on events such as business acquisitions or divorces, research the individual, and prepare a customized email.
The technology itself was useful.
The problem occurred when account managers sent AI-generated emails without sufficiently editing them.
The original outreach sounded different from the account manager’s normal communication style.
When prospects responded and began communicating with the real person, they noticed the disconnect.
That inconsistency damaged trust.
The lesson for executives is important:
AI-generated output still requires human judgment when authenticity and customer relationships matter.
Automation can increase speed.
It can also create new problems when employees are encouraged to accept its output without appropriate review.
The Hidden Problem of Shadow AI
Another obstacle can occur when employees successfully use AI but do not want anyone to know.
Dr. Tsipursky described this as shadow AI usage.
An employee may discover a way to complete a task in 20 minutes that previously required two hours.
Instead of sharing the method with coworkers or management, the employee keeps it private.
Why?
They may fear being judged for using AI.
They may worry management will simply assign them more work.
They may believe admitting AI helped them will diminish the perceived value of their skills.
They may worry that revealing the productivity gain could eventually threaten their job.
The employee becomes more productive individually, but the organization does not necessarily learn from that discovery.
A process that could benefit an entire department remains hidden.
How CEOs Can Reduce Shadow AI
Dr. Tsipursky recommends several practical leadership behaviors.
First, leaders should model AI usage themselves.
Executives can talk openly about how they use AI to analyze information, draft communications, prepare reports, review financial information, or improve workflows.
Visible leadership behavior helps establish that responsible AI usage is not something employees need to hide.
Second, employees need practical training.
Training should focus on real workflows rather than abstract demonstrations of everything an AI platform can theoretically do.
Third, companies can create internal spaces where employees share effective uses of AI.
A dedicated Teams or Slack channel, for example, can give employees a place to share prompts, workflows, lessons, and productivity improvements.
Leaders can reinforce useful contributions with recognition and, where appropriate, rewards.
The goal is to turn isolated AI discoveries into organizational knowledge.
Why “Here Are Your AI Licenses. Go Use Them.” Can Fail
One of the mistakes Dr. Tsipursky sees from executives is deceptively simple.
Leadership purchases licenses to an AI platform, distributes them across the company, and tells employees to start using AI.
The CEO may be enthusiastic because they have already discovered ways AI improves their own productivity.
They assume employees will respond with the same excitement.
Some will.
Others may see an entirely different message:
“Leadership expects me to use this system to automate my own work.”
Technology access is not an adoption strategy.
Leaders still need to address questions of trust, job security, training, professional identity, expectations, quality control, and acceptable usage.
What Is an AI Agent?
The conversation also explored one of the fastest-growing concepts in artificial intelligence: AI agents.
Dr. Tsipursky offered a useful analogy.
Think of an AI system as a newly hired employee who is extremely fast and capable but knows nothing about your company.
You need to provide context, instructions, guardrails, and expectations.
A traditional AI prompt tells that system to perform a task.
Prompt engineering improves the instructions so the AI performs the task more reliably.
An AI agent takes the concept further.
An agent can follow instructions, access information from different sources, perform a sequence of actions, and complete parts of a workflow with less human intervention.
What AI Agents Could Do Inside a Business
Dr. Tsipursky gave an example involving meeting preparation.
Imagine an employee has an upcoming meeting with a prospect.
An AI agent could potentially:
- Review the calendar.
- Identify the meeting and participants.
- Pull relevant information from the CRM.
- Review previous email exchanges.
- Search approved internal company files.
- Gather relevant background information.
- Prepare a meeting briefing.
- Deliver it to the employee before the meeting.
The employee arrives better prepared without manually searching multiple systems.
Similar agents could assist with prospecting, follow-up communications, sales handoffs, performance management, reporting, research, and other recurring workflows.
The Future of Management May Be Managing People Who Manage AI
This raises a much bigger leadership question.
What happens when employees are no longer simply completing tasks themselves?
They may increasingly supervise AI systems that perform parts of their work.
Dr. Tsipursky believes an important management skill will become:
Managing people who manage AI agents.
Consider a sales organization.
A salesperson might manage agents responsible for research, prospect identification, meeting preparation, and follow-up drafts.
An account manager might have a different collection of agents supporting client service and retention.
A manager will still need to lead the humans.
They will also need to understand how those employees’ AI-assisted workflows interact.
Questions surrounding handoffs, accountability, quality, communication, and collaboration do not disappear.
They may become more complicated.
AI May Make Better Human Communication More Important
One of the unexpected topics in our conversation involved prompt engineering.
Learning how to communicate effectively with AI may actually help leaders communicate more clearly with people.
A vague AI prompt often produces a vague or disappointing result.
Better prompts clearly define:
- The desired outcome
- Relevant context
- Specific instructions
- Constraints
- Expectations
- Required format
Those same principles can improve human leadership.
Employees cannot read a manager’s mind.
Vendors cannot automatically infer unstated expectations.
Clearer instructions reduce confusion.
AI may encourage managers to become more deliberate about explaining exactly what they want, how they want it done, and what success looks like.
Humans also benefit from something AI does not necessarily require in the same way:
Why the work matters.
That human context remains an essential part of leadership.
AI Does Not Eliminate the Need for Human Leadership
There is a temptation to frame the future of work as humans versus artificial intelligence.
The more useful question may be how humans and AI work together.
An AI system can draft a customer email.
A human may need to make sure it sounds authentic.
An AI agent can gather information.
A manager may need to decide whether that information supports the right decision.
An automated process can increase speed.
A leader still needs to determine whether speed is improving the customer experience or simply creating mistakes faster.
Artificial intelligence changes the tools available to leaders.
It does not eliminate the need for judgment, communication, accountability, trust, and leadership.
The Most Important Lesson for Business Leaders
Near the end of our conversation, Dr. Tsipursky summarized his message:
It is not the technology. It is the psychology.
That idea deserves attention from any CEO or business owner currently investing in AI.
The technical capabilities of artificial intelligence will continue to evolve.
The business outcome will still depend heavily on how people respond to those capabilities.
Executives need to think about:
- Employee fear
- Professional identity
- Trust
- Training
- Leadership communication
- AI governance
- Customer experience
- Human review
- Knowledge sharing
- Productivity incentives
- AI agents
- Future management skills
The company that buys the most sophisticated AI tool does not automatically gain the greatest advantage.
The more important question is whether leadership can create an environment where people learn how to use AI responsibly, intelligently, and productively.
That requires more than software.
It requires leadership.
Watch or Listen to the Full Conversation
Want to hear Dr. Gleb Tsipursky explain the psychology behind AI adoption, employee resistance, shadow AI, prompt engineering, AI agents, and the future of management?
Listen to the full podcast:
https://open.acast.com/public/streams/5cd334e4e3b953af742edd5d/episodes/6a7e36d56e5b5bfda6370be3.mp3
Watch the full conversation on YouTube:
https://youtu.be/DwrygS2SWj0
Connect With Workforce Alchemy
Website: https://workforcealchemy.com/
Facebook: https://www.facebook.com/ReverseRiskConsulting
Instagram: https://www.instagram.com/workforcealchemy/
X / Twitter: https://x.com/WorkAlchemist
Rumble: https://rumble.com/user/WorkforceAlchemy
Dailymotion: https://www.dailymotion.com/WorkforceAlchemy
YouTube: https://www.youtube.com/@WorkforceAlchemist
Connect with Mason Duchatschek
Website: https://masonduchatschek.com/
LinkedIn: https://www.linkedin.com/in/masonduchatschek/
Subscribe to Workforce Alchemy Magazine
https://workforcealchemymagazine.com/
Subscribe to The Mason Duchatschek Show
https://shows.acast.com/themasonduchatschekshow
Leave A Comment