Overwhelmed? Tick. Confused? Tick. Saturated? Tick. Exhausted? Tick. Disillusioned? Big tick. If you identify with any (or all) of these sentiments when it comes to artificial intelligence (AI), don’t worry, because you’re certainly not alone.
Indeed, experts say AI fatigue is becoming a real reality, and many employees and businesses already seem to be suffering from it.
According to a study by digital adoption platform WalkMe, about 71% of office workers believe new AI tools appear faster than they can learn how to use them. The Opinium study of more than 1,200 office-based professionals also found that nearly half (47%) felt they should be excited about using AI, but instead reported feelings of worry.
According to David Midgley, founder of Squared.io, one of the reasons for this fatigue with AI could be the fact that there is too much talk about it and not enough action.
“The challenge we face is the oversaturation of the discourse about the ‘potential’ of AI, which makes it almost impossible to distinguish between hype and reality,” he says.
“Those investing heavily in AI are overpromising, leaving many companies feeling like their AI partners aren’t delivering. We need to stop talking about the potential of AI and start showing the results of what it can actually achieve today.”
The problem of AI fatigue is inevitable, but also to be expected, according to Dr Clare Walsh, director of education at the Institute of Analytics (IoA).
“For those who have been in digital long enough, they know that there is always a period after the initial excitement of launching a new technology where ordinary users start to realize the costs and limitations of the latest technologies,” she says.
“After 10 years of exciting, uninterrupted advancements – from the first neural networks in 2016 to RAG solutions today – we may have forgotten that this phase of disappointment was coming. This doesn’t negate the potential of AI technology – it’s just an inevitable part of the adoption curve.”
Work fears give way to AI fatigue
With seemingly constant predictions that AI will be responsible for many job losses, it could be that AI fatigue goes hand in hand with this fear among workers and their bosses.
A survey by digital transformation company ArvartoConnect found that a quarter (26%) of UK contact center agents out of a sample of 1,000 were considering leaving their role due to AI anxiety. However, this was not due to the automation itself; it was because of poor communication, low support and unclear career paths.
“Our research reveals the emotional impact that AI deployments have on the very people they are intended to help,” says Debra Maxwell, CEO of ArvatoConnect.
“But this isn’t about stopping innovation; it’s a call to lead with empathy. The organizations that thrive will be those that build trust, not just capabilities. Leaders have a real opportunity to transform uncertainty into accountability.”
Holding back the wave of weariness towards AI also means not presenting it as the only solution to all problems, warns Claus Jepsen, CTO of Unit4.
“It is absolutely essential that the IT team asks the right questions and thoroughly interrogates the business brief,” he explains.
“Very often, AI is not the right answer. If you impose AI on the business when they don’t want or need it, you will experience backlash. You can avoid the threat of AI fatigue if you listen carefully to your team and truly appreciate how they want to interact with the technology, where its use can be improved, and where it adds absolutely no value.”
Research by Opinium from optimization company ABBYY also supports this idea. He spoke to 1,200 or more senior executives at companies with more than 100 employees in Australia, France, Germany, Singapore, the United Kingdom and the United States. Nearly a third (31%) admitted to finding training GenAI models more difficult than expected.
At a time when boardrooms are brimming with ideas for AI implementation, hesitation and fatigue could set in in investment plans. Indeed, the majority of respondents said AI budgets would only increase by 16-20% next year. Only one in ten (11%) said the increase would be 50% or more.
A lack of balance
Corey Keyser, head of AI at Ataccama, a data-trusted software provider, talks about the high expectations around AI: “The problem isn’t that AI is being adopted too quickly; it’s that it’s being adopted without balance. To avoid fatigue, businesses need a barbell strategy.”
“On the one hand, employee adoption. AI should not be imposed in workflows, but made available, accessible and encouraged in a way that feels natural. On the other hand, companies should invest in a small number of AI projects targeted and supported by their leaders,” he adds.
“These initiatives should be directly tied to the organization’s KPIs and built by dedicated teams or consultants who can ensure they are robust, trustworthy, and scalable. Focus on a handful of initiatives with clear ROI and management alignment. This makes adoption targeted rather than performative.”
Speed without structure creates tension, according to Oana Beattie, vice president of data and AI at IT infrastructure company Kyndryl.
“AI fatigue is not just a productivity problem; it’s a risk at the board level,” she says.
“When workflows are disrupted or systems overlap, trust in technology erodes, leading to disengagement, errors, and higher attrition. AI fatigue manifests itself as decreased curiosity and decision fatigue. Left unchecked, it can undermine belonging, morale, and ultimately innovation.”
So-called initiative overload is a major problem that will most likely manifest itself first among workers before moving up the hierarchy to management. But Gavin Guinane, Head of Solutions Engineering EMEA at AI Platform Glean remains – unsurprisingly – optimistic about the current situation.
“We haven’t reached a saturation point for AI, but we have reached a saturation point for fragmentation,” he says.
“The challenge businesses face today is not an overabundance of AI itself, but rather a proliferation of disparate tools that fail to connect across systems, teams and workflows.
“When each department adopts its own standalone solution, organizations find themselves facing a hidden ‘AI tax’ due to duplicated efforts, misaligned data and increasing security risk.
This is where fatigue sets in – not because of AI, but because of the friction of managing disconnected, non-scalable experiences.
Perfect is the enemy of good
Raising the level of AI culture among staff is one way to avoid AI fatigue and contribute to decision paralysis around technology implementation.
This is particularly important given the large number of AI tools to choose from, many of which have similar functions and features, which can lead to duplication and confusion.
Ensuring businesses are aware of evolving AI regulations – often different from country to country – is another key aspect to consider if businesses want to keep their workforce engaged and interested.
There are three common traps that companies fall into, according to John Thompson, senior vice president at consulting firm The Hackett Group and professor at the University of Michigan.
“Waiting to find the perfect use case, not knowing where to start or how to proceed, and taking technological progress as a reason to wait until the level of activity slows down,” he says.
“None of these reasons are good reasons not to move forward, but they are often cited as reasons why progress cannot begin. Senior management should define a mandate by which the organization will begin its AI journey and set broad goals and objectives that can be achieved and measured. IT can and should define the standards, frameworks and architecture that the business operates,” adds Thompson, who is also the author of “The path to artificial general intelligence.’
He further cautions: “What happens when the company spends a lot of money on AI transformation, but it interrupts workflows and productivity instead of contributing to that transformation? It becomes very expensive.”
There is certainly room for improvement among the examples of AI fatigue. Sally Winston, director of XM strategy at Qualtrics – which has a human sentiment database of more than 18,000 customers – points out that this is a “natural evolution rather than true AI saturation”.
She says: “The market has moved from the initial ‘shiny toy’ phase, where organizations rushed to experiment with each new AI tool, to a more mature stage requiring better alignment with business strategy.
“A Qualtrics study shows that while 45% of employees use AI every week, nearly half receive no formal training or guidance. This lack of empowerment, coupled with leaders overestimating employee confidence, leads to employee optimism about AI’s potential that quickly turns into burnout.”
Winston advises leaders to stop saying “yes” to every AI request and better connect AI initiatives to direct business outcomes. “The solution isn’t necessarily fewer AI tools,” she adds, “it’s a better AI strategy. AI will achieve the best results when it meets clear goals with appropriate support.”