A diverse team of colleagues collaborating around a laptop in a Singapore workplace

Workplace Resilience

Leading Across Differences in an AI-Driven Workplace

Resilience Institute · Weekly Resilience Brief · 6 minute read

Aperian’s own research, published by co-founder Dr Ernest Gundling, found culture shapes how AI gets adopted just as much as AI shapes culture in return, particularly around bias, communication style, and leadership dynamics.
A large 2025 study across 18 societies and 2,451 employees found clan-oriented, collaborative cultures generate significantly higher psychological safety than hierarchical ones, with trust in leadership as the key mechanism.
A 2026 scoping review of high power distance environments across Asia found the fix is not removing hierarchy. It is using it deliberately, senior leaders visibly inviting dissent rather than pretending rank does not exist.
This exact tension, AI adoption meeting cultural difference, is the subject of Fit for Change, Across Differences, a Singapore gathering for senior HR and L&D leaders this October.

Most corporate AI training assumes a flat, culture-neutral workplace. Same prompt techniques, same rollout deck, same change management playbook, whether the office is in Austin or Ahmedabad. That assumption does not survive contact with the actual research, and for anyone leading a multicultural team in Singapore, the gap between the assumption and the evidence is worth taking seriously.

Table of Contents

Culture Shapes AI Adoption. AI Reshapes Culture Back.

Aperian, the cultural intelligence firm co-hosting Fit for Change with Resilience Institute, has published research directly on this question. Dr Ernest Gundling, Aperian’s co-founder, argues the relationship runs in both directions. Cultural norms shape how a team actually adopts AI, whether people trust an AI recommendation over a manager’s judgement, whether raising a concern about an AI output feels safe to do out loud. And AI use, in turn, reshapes team culture over time, changing communication patterns and who ends up with more influence in a decision.

That second part is easy to miss. A team that starts routing decisions through an AI tool is not just adding a tool. It is quietly changing who gets listened to, and that shift lands differently depending on how hierarchy already works in the room.

Why Hierarchy Changes the Whole Equation

Psychological safety, the shared belief that a team is safe for interpersonal risk taking, has become one of the most cited ideas in organisational research since Amy Edmondson’s original work. What gets discussed far less is that most of the foundational research on it, including Google’s well known internal study, was conducted primarily in Western, relatively low power distance organisational cultures.

A large 2025 study spanning 18 societies and 2,451 employees found clan oriented cultures, built on collaboration and belonging, generate significantly higher psychological safety than hierarchical ones, with trust in top management as the mechanism connecting the two. A separate 2026 scoping review focused specifically on high power distance environments across Asia reached a more actionable conclusion: trying to eliminate hierarchy outright rarely works. The more effective path is using the hierarchy that already exists, senior leaders actively inviting dissent, visibly rewarding people who raise concerns, and modelling their own fallibility, to build safety across rank rather than pretending rank does not exist.

For a Singapore leadership team, or any team spanning Singapore, India, China, and a handful of Western headquarters, this is not an abstract academic point. It means a psychological safety programme copied unchanged from a US or UK headquarters, built on the assumption of low power distance, will tend to quietly underperform here. Not because the science is wrong, but because it was validated somewhere else.

What This Looks Like When AI Enters the Room

Now add AI to that dynamic. In a high power distance culture, an AI tool’s output can carry a strange kind of borrowed authority. If a junior team member already feels uncomfortable contradicting a senior leader directly, they may feel even less comfortable contradicting an AI recommendation the senior leader has already endorsed. The tool does not just sit neutrally in the room. It can quietly reinforce exactly the dynamic that made speaking up hard in the first place.

This cuts the other way too. A junior employee who is fluent with AI tools can suddenly produce work that looks senior, fast, which scrambles the usual signals a hierarchical culture uses to decide whose voice carries weight. Some organisations are finding this genuinely destabilising. Others are finding it opens a door that used to be closed, a way for someone junior to demonstrate judgement without first needing years of tenure to earn the room’s attention.

Practical Moves for Leading Across This

Separate the AI’s output from the senior leader’s endorsement of it. If a leader shares an AI-generated recommendation, say so explicitly, and explicitly invite disagreement with it. Otherwise the tool inherits the leader’s authority by default.

Build a written or anonymous channel alongside open discussion. Where face saving norms make disagreeing with a senior leader costly in the room, a separate channel captures signal that open discussion alone will miss, AI-related or not.

Do not import a global AI governance playbook without checking it against local hierarchy norms. The controls that make sense in a flat, low power distance headquarters may need real adaptation for a Singapore, Jakarta, or Tokyo office to actually work.

Watch who AI is quietly elevating on your team. If junior staff fluent with AI tools are suddenly producing senior looking work, decide deliberately whether your culture will reward that or resist it, rather than letting the answer emerge by accident.

This exact intersection, AI adoption meeting cultural difference across a multinational team, is what Fit for Change, Across Differences is built around. It is a closed-door gathering of senior HR and L&D leaders in Singapore on 16 October, co-hosted by Resilience Institute and Aperian, with Aperian bringing its own cultural intelligence research directly into the room. It is invite-only with limited seats, RSVP by 30 September. Details are on the event page.

Before your next cross-cultural AI rollout, ask:

Did our AI governance guidelines come from a headquarters with different power distance norms than our own team?
Is an AI tool quietly borrowing a senior leader’s authority in our meetings, without anyone naming that out loud?

AI does not arrive in a culturally neutral workplace, because no such workplace exists. It arrives in yours, with your hierarchy, your history, and your unspoken rules about who gets to disagree with whom, and it amplifies whatever it finds.

FAQ

Frequently Asked Questions

Find answers to some of the most common questions about this topic.

Does AI adoption really work differently depending on a team's culture?

Yes, according to research from Aperian and separate academic studies on psychological safety. Cultural norms shape how much people trust an AI recommendation, how comfortable they feel disagreeing with it, and how AI use in turn reshapes communication patterns on the team over time.

What is the practical difference between a low power distance and high power distance culture for AI rollouts?

In lower power distance cultures, disagreeing openly with a leader or a tool's recommendation carries less social cost, so open discussion channels alone tend to surface concerns. In higher power distance cultures, common across parts of Asia, the same openness carries more risk, so structured or anonymous channels alongside open discussion tend to work better.

Should a global company use one AI governance policy across all its regional offices?

The core principles can stay consistent, but the delivery should adapt. A policy validated in a low power distance headquarters may need real adjustment, more explicit invitations to dissent, additional anonymous channels, to work as intended in a higher power distance office.