Blog 08 - Organisational Culture in the Age of AI: Why Technology Transformation Fails Without Trust, Learning and Employee Voice
Organisational Culture in the Age of AI: Why Technology Transformation Fails Without Trust, Learning and Employee Voice
Organisations often describe Artificial Intelligence (AI) as a technology project. They purchase software, introduce new platforms, automate processes and train employees to use new tools. Yet the success or failure of AI adoption may depend just as much on organisational culture as on the technology itself.
Two organisations can introduce exactly the same AI system and experience completely different outcomes. In one organisation, employees may experiment, share knowledge and use AI to improve their work. In another, employees may hide their use of AI, fear replacement, distrust management or avoid the technology entirely.
What Is Organisational Culture?
Schein (2010) describes organisational culture as a pattern of shared basic assumptions learned by a group as it solves problems of external adaptation and internal integration. His model is commonly understood through three levels: artefacts, espoused values and underlying assumptions.
Artefacts are the visible parts of culture—office design, policies, technology, rituals and observable behaviour. Espoused values are what the organisation says it believes, such as innovation, collaboration or customer focus. Underlying assumptions are the deeper beliefs employees may take for granted: whether mistakes are tolerated, whether management can be trusted, whether information is safe to share and whether employees are genuinely valued.
This framework explains why simply introducing AI does not create an innovative culture. An organisation can purchase advanced technology—an artefact—while retaining deeper assumptions that discourage experimentation or punish failure.
The AI Readiness Gap Is Also a Cultural Gap
Microsoft’s 2025 Annual Work Trend Index illustrates how differently leaders and employees may experience the same technological transition. The report found that 67% of leaders were familiar with AI agents compared with 40% of employees. It also found that 79% of leaders believed AI would accelerate their careers compared with 67% of employees (Microsoft, 2025).
Figure 1. Differences in AI readiness between leaders and employees. Author’s illustration based on Microsoft (2025).
This gap matters because leaders may interpret AI as opportunity while employees interpret the same change as uncertainty. If management announces that “AI will transform the business” without explaining what that means for individual roles, employees may reasonably wonder whether transformation is simply another word for job reduction.
Change therefore requires more than executive enthusiasm. Employees need practical understanding of how AI affects their work, what skills they need, which decisions remain human and how the organisation will respond if roles change.
A Culture of Experimentation—Without Fear
The CIPD (2025) argues that organisations can help employees become more comfortable with AI by creating a culture of experimentation and shared learning. This is an important cultural principle because employees cannot develop confidence in a new technology if every mistake is treated as failure.
Edmondson’s (1999) concept of psychological safety is particularly relevant. Psychological safety exists when team members believe they can speak, ask questions, admit mistakes and raise concerns without being humiliated or punished.
AI creates many situations in which employees need exactly this behaviour. They may need to say:
- “I do not understand how this AI system reached its answer.”
- “The output looks wrong.”
- “I used AI for this task and I need someone to check the result.”
- “This system is creating a problem for customers.”
- “I think this automated process may be unfair.”
If employees are afraid to raise those issues, AI risk becomes harder to detect. A culture that looks disciplined may actually become dangerous because employees learn to remain silent.
An organisation cannot demand innovation while punishing every unsuccessful experiment. Employees need clear boundaries for responsible AI use, but they also need enough psychological safety to learn.
Trust Determines How Much Authority Employees Will Give AI
Trust is another important cultural factor. A 2025 CIPD poll of 2,214 respondents found that 63% would trust AI to inform important work decisions, but only 1% would trust AI to make important work decisions (CIPD, 2025).
Figure 2. Trust in AI as decision support compared with AI as decision-maker. Author’s illustration based on CIPD (2025).
This is an important distinction for organisational culture. Employees may welcome a system that helps them analyse information while resisting a system that appears to replace judgement, accountability and empathy.
The CIPD therefore recommends a human-centred approach with clear principles for AI use, including ethical practice, fair treatment and human oversight. Trust is unlikely to develop simply because management tells employees that a system is reliable. It develops when employees understand how technology is used and see that the organisation responds responsibly when problems occur.
Culture Is Created by What Leaders Do, Not Only What They Say
Leaders play an important role in shaping organisational culture because employees observe what leaders pay attention to, reward and tolerate.
A company may publish an AI policy saying that employees should “use AI responsibly”. However, if managers reward only speed and never ask whether AI output is accurate, employees learn that speed matters more than responsibility.
Similarly, an organisation may say it supports learning while expecting employees to develop AI skills entirely outside working hours. The visible message is “learning matters”; the cultural message becomes “learning is your personal problem”.
Employee Voice Can Determine Whether AI Is Accepted
Employees are more likely to understand and accept technological change when they are involved in it. Worker consultation also gives organisations access to practical knowledge that senior managers and software developers may not possess.
The OECD’s 2025 study of Japan illustrates how much consultation differs across contexts. In Japan, only 16.3% of employees in finance and insurance reported being consulted when new technologies were introduced, compared with 49.1% across seven surveyed OECD countries. In manufacturing, the figures were 16.4% in Japan compared with 42.4% across the seven-country average (OECD, 2025).
Figure 3. Employee consultation during technological change in Japan compared with seven surveyed OECD countries. Author’s illustration based on OECD (2025).
The same OECD report states that a lack of trust can increase worker resistance and hinder AI adoption. This supports the argument that employee voice is not simply a “soft” HR practice. It can affect whether technological transformation works.
Employees should therefore be able to influence questions such as:
- Which tasks should AI support?
- Which tasks should remain primarily human?
- What employee data should be collected?
- How should AI-generated errors be reported?
- What training is required?
- How should employees challenge automated decisions?
Learning Culture Versus Tool Culture
Organisations sometimes mistake access to tools for capability. Purchasing an AI licence for employees does not mean employees know how to use it safely or effectively.
A genuine learning culture provides employees with time, examples, peer support, training and opportunities to practise. Employees also need to learn when AI should not be used and how to verify its output.
The CIPD (2026) notes that AI can improve working lives when it is used to automate repetitive tasks and free employees for higher-value or more enjoyable work. However, achieving that outcome requires deliberate work design. If every saved minute is simply converted into a higher target, employees may experience AI as work intensification rather than support.
AI Is Also Changing Organisational Structure
AI adoption may eventually change not only individual jobs but also how organisations are structured. Microsoft’s 2025 Work Trend Index reports that 81% of surveyed leaders expected AI agents to be moderately or extensively integrated into their organisation’s AI strategy within the following 12–18 months. Leaders also expected employees increasingly to train and manage agents.
This raises cultural questions about authority and expertise. If a junior employee using AI can complete analytical work previously associated with more senior roles, organisations may need to reconsider how status and career progression are defined. Expertise may become less about personally producing every output and more about framing problems, checking quality, applying judgement and coordinating human and digital resources.
Change Management: From Announcement to Participation
AI transformation is also a change-management challenge. Kotter (1996) argues that organisational change requires a clear vision, leadership, communication, broad participation and reinforcement of new behaviours.
This suggests that simply launching a new tool and providing one training session is not enough. Organisations need to explain why the change is happening, identify credible use cases, support employees through experimentation and adjust policies as the organisation learns.
A healthy AI culture should combine:
Purpose — employees understand why AI is being introduced.
Psychological safety — people can question outputs and admit mistakes.
Trust — significant decisions retain human accountability.
Learning — employees receive time and support to build capability.
Employee voice — affected workers influence implementation.
Clear boundaries — responsible use, privacy and security expectations are understood.
Leadership example — managers model the behaviours they expect from others.
A Global HRM Perspective
Organisational culture also interacts with national culture and institutions. Brewster et al. (2017) emphasise that International HRM must account for differences in legal systems, employment relations, social expectations and managerial traditions.
The OECD findings from Japan are a good example. Lower levels of worker consultation and lower trust in employer use of safe and trustworthy AI do not simply describe the technology. They reflect the wider relationship between organisations, employees and institutional practices.
A multinational organisation therefore cannot assume that one AI-change programme will create the same response in every country. A highly participative approach may feel familiar in one context but unusual in another. Privacy expectations, employee voice mechanisms and attitudes towards authority can also vary considerably.
Global HRM should therefore seek consistent ethical principles while adapting communication, consultation and implementation to local contexts.
🎥 Recommended Video
This CIPD discussion connects AI transformation with employee empowerment and organisational change readiness, making it directly relevant to the cultural issues explored in this post.
▶ Watch on YouTubeMy Reflection
Before exploring this topic, I mainly thought that successful AI adoption depended on choosing the right technology and providing enough training. I now believe those are only part of the challenge.
Employees also need to trust the organisation’s intentions. They need to know whether AI is being introduced to support their work, monitor them more closely or eventually reduce roles. If management avoids those questions, employees will create their own explanations.
The difference between the CIPD findings on trusting AI to inform decisions and trusting AI to make decisions was particularly interesting to me. It demonstrates that people may accept technology while still wanting human accountability. Therefore, resistance to full automation should not automatically be interpreted as resistance to innovation.
I also see employee voice differently after examining the OECD evidence. Consultation is not simply about asking employees whether they “like” a new system. Employees can identify practical risks, workflow problems and unintended consequences that decision-makers may miss.
For me, this means AI transformation should be treated as a cultural and people-management process, not simply a software implementation.
💬 Join the Discussion
What matters more for successful AI adoption: having the best technology or having the right organisational culture?
Would you trust an organisation more if employees were involved in deciding how AI is used? What behaviours should leaders demonstrate if they want employees to experiment with AI responsibly? Share your view in the comments.
References
Brewster, C., Sparrow, P., Vernon, G. and Houldsworth, E. (2017) International Human Resource Management. 4th edn. London: CIPD.
CIPD (2025) Almost two thirds of people would trust AI to inform important work decisions, CIPD poll shows. 28 January. Available at: https://www.cipd.org/en/about/press-releases/almost-two-thirds-people-trust-ai-to-inform-important-work-decisions/ (Accessed: 12 August 2026).
CIPD (2026) AI and technology. Available at: https://www.cipd.org/uk/knowledge/factsheets/ai-technology/ (Accessed: 12 August 2026).
Edmondson, A. (1999) ‘Psychological safety and learning behavior in work teams’, Administrative Science Quarterly, 44(2), pp. 350–383.
Kotter, J.P. (1996) Leading Change. Boston, MA: Harvard Business School Press.
Microsoft (2025) The 2025 Annual Work Trend Index: The Frontier Firm is born. 23 April. Available at: https://blogs.microsoft.com/blog/2025/04/23/the-2025-annual-work-trend-index-the-frontier-firm-is-born/ (Accessed: 12 August 2026).
OECD (2025) Artificial Intelligence and the Labour Market in Japan. Paris: OECD Publishing. Available at: OECD report (Accessed: 12 August 2026).
Schein, E.H. (2010) Organizational Culture and Leadership. 4th edn. San Francisco, CA: Jossey-Bass.
Clegg, S., Courpasson, D. and Phillips, N. (2006) Power and Organizations. Newbury Park, CA: Pine Forge Press.
I strongly agree that technology transformation can fail if employees do not trust the organisation introducing it. Employees may resist AI not necessarily because they dislike technology, but because they fear job loss, increased monitoring or losing control over their work. This means organisational culture and employee voice become important during digital transformation. In my view, leaders need to communicate honestly about why AI is being introduced and involve employees in the implementation process. Technology transformation should therefore be managed as a people change process rather than simply an IT project.
ReplyDelete