Part 1 — The leadership challenge hiding behind our conversations about technology, governance and risk.
The Question That Keeps Coming Back
We talk a great deal about AI. You really can’t get away from it these days.
Its potential to transform productivity, reshape industries, improve decision-making and fundamentally change how organisations operate is seemingly everywhere. We’re hungry for case studies, exploring emerging technologies, building governance frameworks and trying to keep pace with evolving regulatory expectations and the risks accompanying increasingly autonomous systems.
And rightly so.
There is enormous potential in what we’re exploring, alongside some equally significant challenges.
But there’s a question that keeps resurfacing in these conversations, particularly among those responsible for leading and governing organisations.
Who is really accountable?
It’s a reasonable question. Yet I wonder whether we’re spending enough time examining what we’re actually asking.
Because while AI may be introducing new complexities into our organisations, the challenges surrounding responsibility, accountability and ownership are far from new.
Perhaps AI is simply making them harder to ignore.
And perhaps the more revealing question isn’t whether we’ve established accountability, but whether we genuinely understand what it means to exercise it.
A Conversation That Stayed With Me
I recently attended the Australian Institute of Company Directors (AICD) Essential Director Update on the Gold Coast, hosted by Greg Ewing, AICD’s Queensland State Manager.
It was an excellent opportunity to hear different perspectives on the challenges facing boards, including AI transformation, organisational culture, productivity, resilience, ethics and the evolving expectations of directors.
Penny Bingham-Hall, Non-Executive Director at Fortescue, shared insights into the organisation’s approach to AI, including Alfie, an AI-enabled planning capability designed to navigate complex operational decisions across its supply chain.
What stood out wasn’t simply the technology.
It was the broader approach presented: aligning AI with business strategy, involving employees in redesigning workflows, considering workforce impacts and recognising that board oversight must extend beyond technology and cybersecurity to encompass people, culture, sustainability and organisational performance.
For a capital-intensive business, the commercial rationale for improving efficiency and productivity is clear. But the presentation also recognised that technological transformation has a human dimension.
It’s an encouraging approach, and a valuable example of the thinking boards can bring to AI transformation.
Of course, a well-designed strategy, thoughtful governance arrangements and good intentions are only the beginning. Their real value emerges through implementation, particularly in how people are engaged, how decisions are experienced and whether the intended benefits are realised.
A transformation may be strategically sound and commercially compelling, but that doesn’t automatically mean everyone affected will understand, support or experience it in the same way.
And that matters.
Simon Mitchell, AICD’s Acting Head of Policy, then explored the wider operating environment confronting Australian boards: declining productivity, regulatory complexity, supply chain resilience, cyber threats, ethics and increasingly demanding expectations of oversight.
Simon also shared an observation attributed to Jane McAloon, Chair of BlueScope, that captured the challenge particularly well:
"Governance is not waiting for certainty. It is making sound decisions before certainty arrives."
It’s a powerful reminder that governance doesn’t offer the luxury of waiting until every uncertainty has been resolved.
Directors must still exercise judgement, consider consequences and remain accountable for the decisions they make.
And, of course, we don’t all hear or interpret the same information in the same way. We arrive with different experiences, questions and assumptions, sometimes looking for reassurance, sometimes seeking answers and sometimes hoping to challenge an established perspective.
That’s part of the value of bringing people together for conversations like these.
The presentations and subsequent panel discussion offered valuable perspectives on how organisations and boards are navigating an increasingly complex environment.
But it was some of the conversations outside the formal sessions that stayed with me.
Despite the wealth of information being shared, several AICD members expressed a desire for greater depth around what accountability actually means for boards in the context of AI.
One member raised a deceptively simple example.
If an AI chatbot produces a response that creates a problem, who is accountable?
And there it was again.
That familiar question.
Not necessarily because accountability had been overlooked in the presentations, but because discussing governance and accountability as principles doesn’t always resolve the questions people face when trying to apply them.
And perhaps that’s where we need to look more closely.
Are We Asking the Right Question About Accountability?
When something goes wrong, our instinct is often to establish why it happened and, of course, who was responsible.
Who approved the technology? Who implemented it? Who was overseeing it? Who should have identified the problem? Who should have intervened?
These are legitimate questions.
Though are we beginning the conversation too late?
Because accountability doesn’t suddenly appear when an adverse outcome occurs.
It exists in the decisions made long before that outcome becomes visible.
The decision to adopt the technology.
The assumptions used to justify its implementation.
The authority granted to individuals, teams and systems.
The safeguards established, or overlooked.
The information considered, or dismissed.
The concerns raised, or ignored.
And the decisions to act, delay, challenge or do nothing.
In the chatbot example, accountability may sit across different roles and decisions. A board has responsibilities for governance, oversight and, increasingly, strategy. Executives are accountable for decisions within their authority. Teams and individuals have responsibilities for how systems are designed, deployed, monitored and used.
The precise allocation depends on the circumstances, including governance arrangements, delegated authority and applicable legal obligations.
And we shouldn’t confuse accountability for the governance of a system with responsibility for every individual action or output that system produces.
But the underlying principle remains.
The introduction of technology does not remove human accountability for the decisions surrounding its use.
Nor should we assume that because a governance structure identifies who is accountable, everyone involved necessarily understands what that means in practice.
Perhaps this is another dimension of what I’ve previously described as the AiLlusion: the belief that increasingly sophisticated technology and governance arrangements necessarily reflect an equally sophisticated understanding of the human responsibilities surrounding them.
Yet accountability isn’t established simply because we’ve documented who is responsible. It’s demonstrated through the decisions we make, the judgement we exercise and our willingness to own the consequences of our actions and inactions.
And this raises a more uncomfortable question.
Do we genuinely understand how accountability operates throughout our organisations, or have we become so accustomed to assigning it that we’ve stopped examining what it means to exercise it?
Three Dimensions We Too Often Treat as One
Responsibility. Accountability. Ownership.
Three words frequently used in leadership conversations, governance frameworks, position descriptions and performance expectations.
Sometimes almost interchangeably.
Yet each brings a different dimension to how people understand their role, their decisions and their relationship with organisational outcomes.
It’s tempting to assume everyone understands these distinctions. After all, they’re familiar words, embedded in our governance frameworks, leadership expectations and everyday conversations.
But understanding something in principle and understanding what it means in practice are not necessarily the same thing.
And perhaps the more revealing question is whether leaders themselves share the same understanding.
Because the expectations established in the boardroom don’t always translate into the same understanding throughout an organisation.
And as I mentioned earlier, we can all view the same information and interpret it differently. Therein lies a risk in assuming we all interpret it the same way.
Responsibility: What Am I Expected to Do?
Responsibility concerns the duties, decisions and actions associated with a person’s role.
What am I expected to deliver?
What decisions fall within my authority?
What obligations have I accepted?
And where do the boundaries of my role begin and end?
Responsibilities can be defined, allocated and delegated.
But clarity matters.
When expectations are ambiguous, authority is poorly understood or responsibilities overlap without coordination, people may act on different assumptions about what is expected of them.
And those assumptions can create risk long before anything visibly goes wrong.
Consider the introduction of an AI capability into an existing business process.
The technology team may be responsible for configuring the system. The business unit may be responsible for the operational process. Risk, legal and compliance teams may provide advice or oversight within their respective functions.
Each may be fulfilling its assigned responsibilities.
But who is responsible for ensuring that the decisions made across those functions come together into a coherent and appropriately governed outcome?
That question may have a clear answer in a well-designed governance arrangement.
But have we tested whether everyone involved shares that understanding?
Because responsibility isn’t simply about completing the tasks allocated to us.
It’s also about understanding how our actions, decisions and dependencies connect with those of others.
Accountability: What Must I Answer For?
Accountability extends beyond completing an assigned task.
It involves being answerable for decisions, actions, inactions and their consequences within the authority and obligations associated with one’s role.
It also requires a willingness to explain the reasoning behind decisions, acknowledge mistakes, address consequences and learn from experience.
But accountability is not simply about accepting blame when something fails.
Nor is it only relevant when a result is negative.
Accountability is exercised through the decisions we make every day, including the decisions we choose not to make.
And it must exist at every level.
A board cannot simply delegate away its own governance obligations. Executives cannot avoid accountability for the direction they establish. Equally, individuals cannot disregard their own responsibilities or behaviours simply because a decision originated elsewhere.
The nature and extent of accountability may differ, but the principle does not disappear.
And this is where I think we need to challenge a common assumption.
Assigning accountability is not the same as ensuring it is understood, accepted or exercised effectively.
An organisation may have clearly documented reporting lines, delegations of authority, performance expectations and governance structures.
But what happens when the person expected to exercise accountability doesn’t believe they have the authority to influence the outcome?
Or when someone has the authority but lacks the knowledge, judgement or willingness to exercise it appropriately?
What happens when a decision crosses multiple functions, each with different priorities and interpretations of accountability?
And what happens when people are reluctant to challenge decisions because they fear the consequences?
These questions don’t diminish individual accountability.
They remind us that accountability operates within a broader organisational environment, one that leaders help create and influence.
And perhaps that environment deserves just as much scrutiny as the accountability structures themselves.
Ownership: Do I Genuinely Care About the Outcome?
Ownership introduces another dimension.
It is the degree to which someone genuinely commits to the purpose, decisions and outcomes associated with their work.
Not simply because they have been told to.
Not merely because it appears in a position description.
But because they recognise their contribution, understand its significance and choose to engage with the outcome.
And this is where things become interesting.
We can assign responsibility. We can establish accountability. But genuine ownership cannot simply be demanded into existence.
We can expect people to meet their obligations, regardless of whether they personally agree with every decision.
But if we want them to contribute their judgement, identify emerging risks, challenge assumptions and actively pursue better outcomes, we need to consider the conditions in which they are expected to operate. That doesn’t mean everyone must have a say in every decision. Nor does it mean people are relieved of their obligations when they disagree with the direction being taken.
But there is a difference between expecting someone to fulfil their responsibilities and creating an environment in which they genuinely want to contribute to the outcome. And that distinction becomes particularly important during organisational transformation.
We can communicate a strategy, explain the intended benefits, provide training and establish implementation plans. However, have we genuinely engaged the people whose knowledge, experience and judgement may determine whether the transformation succeeds?
Have we listened to what they understand about the existing processes, the risks and the unintended consequences?
And when they raise concerns, are those concerns treated as valuable insights or obstacles to progress?
Because genuine ownership is not simply about enthusiasm for change.
It’s about understanding the purpose, recognising one’s contribution and being willing to exercise judgement in pursuit of a meaningful outcome.
Which raises another uncomfortable question.
Are we creating environments that encourage genuine ownership, or are we simply expecting people to take ownership of decisions they had little opportunity to understand or influence?
The Accountability Question We Rarely Ask Ourselves
This brings me back to those conversations at the AICD event.
When directors and executives ask who is accountable for an AI-generated response, it’s an important question.
But I wonder whether there’s an even more fundamental conversation we need to have.
What does accountability mean to us personally?
How do we interpret it?
How do we respond when we’re expected to exercise it?
And perhaps most revealingly, how do we respond when someone else holds us accountable?
Because accountability is not simply a concept we establish through governance structures. It’s something we experience through our relationships, decisions and interactions with others.
For some, accountability represents trust, responsibility and the opportunity to make a meaningful contribution.
For others, the same word may trigger uncertainty, defensiveness or fear.
And sometimes, accountability can become something far less constructive.
A means of shifting blame, protecting authority, silencing challenge or demanding from others what we are unwilling to demonstrate ourselves.
When accountability becomes a weapon, its purpose can be distorted from strengthening responsibility and trust to reinforcing fear and self-protection.
But we need to be careful here. Holding someone accountable is not inherently punitive. Nor should accountability be avoided simply because it creates discomfort.
There are times when difficult conversations, consequences and corrective action are entirely appropriate.
The distinction lies in how accountability is understood and exercised, whether expectations are fair, whether authority and circumstances are considered, and whether the same standards apply to those establishing the direction.
And perhaps that’s where leadership must be willing to look in the mirror.
Because it’s one thing to ask whether the people in our organisations understand accountability.
It’s another to ask what our own behaviour has taught them about it.
Before We Ask Who Is Accountable for AI...
AI is creating extraordinary opportunities, alongside new challenges in governance, authority, risk and decision-making.
We need appropriate frameworks, controls, oversight and assurance.
And we need boards and executives who understand the decisions they’re making, the authority they’re delegating and the consequences they remain accountable for.
But technology cannot resolve a lack of shared understanding about responsibility, accountability and ownership.
Nor can governance structures, however sophisticated, guarantee that people will exercise sound judgement, challenge assumptions or willingly embrace the outcomes they’re expected to deliver.
Those are fundamentally human dimensions of leadership.
And they existed long before AI entered our organisations.
Perhaps that’s why the accountability question keeps returning.
Not because we lack terminology, policies or governance structures, but because understanding accountability requires us to look beyond what is written and examine how it is actually experienced and exercised.
So, before we become too comfortable asking who is accountable when AI gets something wrong, perhaps we should be asking:
Do we genuinely understand responsibility, accountability and ownership throughout our organisations?
Have we created the conditions for people to exercise them effectively?
And are we prepared to hold ourselves to the same expectations we establish for others?
Because risk isn’t simply something to be managed through systems, controls and compliance.
It’s a leadership lens.
And sometimes the most important risk we need to confront is the one looking back at us in the mirror.
What say you?
In Part 2, we’ll explore why accountability can inspire commitment in some people, provoke fear in others and sometimes become a weapon. We’ll look beneath the expectations and governance structures to examine the human behaviours, organisational conditions and leadership influences that shape our relationship with accountability.
