AI questions weaken workplace learning ties

Employees are increasingly asking AI tools questions they would previously have taken to colleagues, reshaping informal learning, team relationships, and the practical value of in person work.

A MyIQ analysis of 22,481 adults found that 74% of regular AI users now ask the technology questions once directed to co-workers. Around 48% reported fewer spontaneous conversations during the working day, while 53% said work feels more transactional since adopting AI.

The research points to a subtle but important shift in how knowledge travels through organisations. AI can make workers more independent, but it can also reduce the everyday exchanges that help colleagues build trust, learn judgement, understand context, and become known inside a team.

MyIQ Managing Director Sarah Meyer wrote: “Repeated across a working week, those missing exchanges can mean fewer opportunities to build trust, share judgment, and become known inside a team.”

The data also shows the appeal of AI. Nearly two-thirds of respondents said they feel more comfortable making decisions independently than they did a year ago, while 71% said they feel more self-sufficient at work. The technology is reducing friction in everyday problem solving, particularly where workers need fast answers, drafts, summaries, or explanations.

Many workplace questions, however, are not purely factual. A colleague may answer with organisational context, political judgement, customer knowledge, or a warning that a formal process works differently in practice. AI can provide speed and structure, but it may not know the lived realities of a team unless that context has been captured and governed.

The findings add another layer to the UK AI adoption debate. Analysis in UK AI adoption barely deepens found that adoption has widened faster than it has matured, with many companies using AI in limited ways. The MyIQ research suggests that even relatively shallow use can still change workplace behaviour if employees redirect everyday questions away from colleagues.

It also complicates return to office arguments. Many employers have justified office attendance partly on the basis of spontaneous collaboration, mentoring, and cultural connection. If routine questions are increasingly being handled by AI, offices need purposeful interaction rather than reliance on proximity alone.

Newer employees may be particularly affected. Informal questions are often how junior staff learn norms, judgement, shortcuts, and relationships. If those questions are diverted to a chatbot, onboarding may become more efficient in some respects but weaker in social integration. A worker may complete tasks faster while understanding less about how decisions are made.

Managers will need to separate two issues that are often blurred. The first is whether AI improves individual productivity. The second is whether the organisation is preserving the human networks that allow teams to coordinate, challenge, mentor, and adapt. Efficiency can improve in one area while capability weakens in another.

There is a governance dimension as well. If employees are asking AI tools questions about policies, customers, contracts, data, finance, HR, or regulated work, companies need clarity on which tools are approved, what information can be entered, how answers should be checked, and when escalation to a person is required.

The problem is not solved by discouraging AI use. Workers are turning to these tools because they are quick, available, and often useful. A blanket resistance would ignore the productivity gains many employees are already experiencing. The more realistic task is to design working practices that combine fast machine support with deliberate human learning.

That could mean clearer norms around when to ask colleagues, when to use AI, and when to verify answers. It could also mean more structured mentoring, team check-ins, office days built around collaborative work, and shared knowledge bases that make internal context more accessible.

The technology creates a leadership test that will not always show up in output data. A team may appear more efficient while becoming less connected. The warning signs may be weaker onboarding, fewer cross team relationships, less challenge, slower escalation of risks, or reduced trust.

AI is already part of the daily workflow in many organisations. The next phase will depend on whether companies treat it as a private productivity tool or as a force that changes how knowledge, judgement, and relationships circulate at work.