Deloitte’s disastrous use of generative AI in a government-commissioned report highlights a critical question for leaders: what role does culture play in AI adoption?
Deloitte has since admitted that parts of its $440,000 report for the Albanese Government contained fabricated references generated by AI and has refunded part of its fee. The firm maintains the recommendations were sound, but the reputational damage was immediate and widespread (The Guardian, 6 October 2025).
The issue here is not just a faulty citation. It’s a failure of alignment between technology and organisational culture.
When AI is deployed without the right cultural foundations—transparency, accountability, ethical awareness—trust collapses. And once trust is lost, no amount of technical sophistication can restore it.
Culture: The Invisible Architecture Behind AI Decisions
AI systems don’t exist in a vacuum. They are products of human intention, reflecting the assumptions, incentives, and values of the organizations that create and use them.
A culture that prizes speed and efficiency above all else will use AI to move faster—sometimes at the expense of quality or ethics.
A culture grounded in reflection, integrity, and curiosity will use AI to deepen insight and broaden participation.
This is why AI implementation is never just a technical project. It is a cultural one.
Every AI decision—what data is used, what outcomes are prioritised, how outputs are validated—expresses a cultural choice.
In Deloitte’s case, the technology worked as designed; the failure lay in how it was governed. The cultural systems—oversight, accountability, disclosure—did not keep pace with technological adoption.
Trust Is a Cultural Asset
Trust is earned when people believe that decisions are made transparently, ethically, and with care. AI challenges that trust in two ways:
- Opacity: AI-generated outputs can appear authoritative, even when flawed.
- Accountability gaps: When responsibility is unclear, credibility erodes.
The Deloitte episode shows how quickly credibility can unravel when these cultural guardrails are weak. Leaders cannot rely on disclaimers after the fact. They need cultural disciplines built in from the start—clarity about when AI is used, who reviews its output, and how errors are addressed.
Building a Culture Fit for the Age of AI
To navigate the intersection of AI and culture, organisations need more than ethical codes or AI policies. They need living cultures—environments that shape everyday choices, not just compliance.
Five cultural principles are emerging as essential:
- Transparency – Disclose when and how AI is used. Be open about its limitations.
- Accountability – Keep human oversight at every decision point.
- Ethical literacy – Equip teams to understand how bias, hallucination, and automation can distort outcomes.
- Inclusion – Involve diverse voices in AI design and review to surface unseen risks.
- Learning orientation – Treat missteps as opportunities to refine systems and strengthen governance.
These principles ensure AI serves the organisation’s values—not the other way around.
The Leadership Challenge
The Deloitte case is not an isolated misstep; it’s a signal. As AI becomes more embedded in decision-making, leaders must redefine what good governance looks like.
This is not about banning AI—it’s about embedding it responsibly within a culture that values integrity over expedience. The question for leaders is no longer “Can we use AI?” but “Does our culture make us trustworthy custodians of it?”
Organisations that treat culture as a strategic asset—not a soft concept—will be the ones that turn AI from a reputational risk into a trusted ally.
Final Reflection
AI is a mirror. It reflects the culture of the organization that wields it. If we want trustworthy AI, we must first build trustworthy cultures—cultures that balance ambition with ethics, innovation with introspection, and speed with accountability.
The Deloitte affair is not simply a cautionary tale—it’s an invitation to lead differently. Because in the age of AI, culture is the true operating system.