AI adoption in the workplace has stagnated, with a survey by Gallup revealing that the percentage of workers using AI increased only slightly from 45 percent in the third quarter of 2025 to 46 percent in the fourth quarter. Despite this, the number of employees reporting frequent use of AI has risen, with a two percent increase in daily usage. Knowledge-based industries, especially in the technology sector, lead in AI adoption, with 77 percent of tech professionals using AI at work. Gallup suggests a growing divide between users and non-users, indicating a lack of clear use cases for AI among lower-level employees. The survey highlights the need for leadership to understand how AI can be applied across different roles, not just within decision-making circles.
A recent study by Morgan Stanley reveals that the UK is experiencing greater job losses due to artificial intelligence than other major economies, such as the United States, Japan, Germany, and Australia. The research indicates that British businesses have reported a net job loss of 8% over the past year, despite an 11.5% increase in productivity attributed to AI. Additionally, a survey by Randstad found that more than a quarter of UK workers fear losing their jobs to AI within the next five years, with younger workers expressing the most concern.
Why do we care?
AI isn’t failing to spread because employees are slow. It’s failing to spread because leaders are using it first, fastest, and most aggressively — and not redesigning the rest of the organization to come with them.
When productivity goes up and headcount goes down, workers don’t see AI as a tool. They see it as a signal. Especially younger workers, who are being told they need to adapt while watching roles quietly disappear.
For MSPs, this is where bad advice causes real harm. If you pitch AI as a leadership accelerator without addressing how work changes below that layer, you’re setting customers up for resistance, disengagement, and eventual retrenchment.
AI that only lives in decision-making circles doesn’t transform organizations. It compresses them. Operationally, this is the mistake to avoid. If AI is deployed as a leadership accelerator without redesigning roles, workflows, and failure paths below that layer, you’re not improving capability — you’re removing slack. That creates brittle organizations that look efficient until something breaks.
The immediate temptation is to celebrate productivity numbers. The smarter move is to ask: who actually got more capable? If the answer is “only the top,” then AI isn’t being adopted — it’s being weaponized against the org’s own resilience.