.png)
Should we hire a CDO, a CDAO or a CAIO?
For many organisations, this is no longer a question about creating data or AI leadership from scratch. A Chief Data Officer, Chief Data and Analytics Officer, CIO, CTO, or senior data leader is already in place, and some responsibility for AI already sits somewhere across that group.
What is changing is the scale of the data and AI work. AI is increasing the importance of data while also stretching beyond the traditional boundaries of the data function. Should data, analytics and AI continue to sit together, or has AI become broad enough to require separate executive ownership?
While AI clearly deserves executive attention, does that responsibility naturally belong with the data leader already in place? And if it does, does that person have the scope, authority and capacity to lead it properly?
AI is making data more important, not less
Models and algorithms are only part of what determines whether AI can scale. The quality, accessibility and governance of the underlying data remain critical.
McKinsey reported in 2026 that data was emerging as a major constraint as organisations tried to scale AI, with more than two-thirds of high-performing companies identifying data as the primary obstacle to enabling it.
Lucent Search's own research shows how these dependencies are affecting AI delivery.
91%
of AI leaders struggle to move pilots into production
89%
face challenges integrating AI with legacy systems
The research found that many pilots are being developed without the data pipelines, integration architecture and production environments required to operate at enterprise scale. AI cannot create much value if it cannot connect reliably with the systems and information through which the organisation actually operates.
Creating separate AI leadership does not remove those dependencies. If data quality, access, integration or existing technology remain significant constraints, the relationship between the AI leader, data leader and technology leadership becomes more important rather than less.
The data issues organisations have spent years trying to resolve — ownership, quality, access, lineage, governance and consistency — do not disappear when AI arrives. In many cases, AI simply makes the consequences of not resolving them more visible.
The starting point matters more than the title
The difference between a CDO and CDAO is rarely clean enough to treat the two as standardised jobs.
Some CDOs already lead analytics, data science and AI. Some CDAOs are still heavily focused on governance, ownership and fundamental data quality. In other organisations, the data executive has a much broader role around insight, products and business performance.
The more useful starting point is what the organisation already needs that executive to achieve.
Deloitte's 2025 UK Chief Data Officer research illustrates this well. Organisations reporting lower levels of data maturity were more likely to focus on data governance and data strategy, while those reporting higher maturity placed more emphasis on AI, generative AI, and data products.
The role evolves according to what is already working well and where the organisation still has work to do, rather than following a progression from CDO to CDAO and then to CAIO.
An organisation with significant problems around data ownership, quality and governance may need its senior data leader to remain heavily focused on those issues even as AI investment increases. An organisation with stronger foundations may reasonably expect the same executive to spend more time on analytics, AI and the value created from the data estate.
Quite often, what appears to be an AI problem is at least partly a data problem. IBM's 2025 research with 1,700 CDOs found that only 26% were confident their data capabilities could support new AI-enabled revenue streams, with accessibility, completeness, integrity, accuracy and consistency among the barriers identified.
If data readiness is constraining AI adoption, separating the leadership may create more complexity rather than less. A new CAIO cannot make an underlying data problem disappear, so responsibility for resolving it needs to be clear across the data and AI leadership.
Without that clarity, two senior executives can end up sharing accountability for an outcome that depends on both of them.
When AI can sit naturally with the data leader
There is a strong case for keeping data, analytics and AI together when the AI agenda is closely connected to capabilities the data function already owns.
Keeping the responsibilities together is particularly credible where AI is developing through analytics, data science, machine learning and data products, and where the existing data leader has strong relationships across the business and enough influence to shape how data and AI are used.
Gartner's 2025 survey of 504 data and analytics executive leaders found that 70% of CDAOs had primary responsibility for AI strategy and operating model.
The executive responsible for AI remains close to the data it depends on. Analytics, data science and AI capability can be developed together. The organisation also avoids introducing another senior boundary across areas that are already highly interdependent.
Governance reinforces that connection, although data governance and AI governance are not the same thing. Data leaders are already dealing with ownership, quality, lineage, access and appropriate use. AI adds further questions around models, use cases, risk, accountability and responsible deployment.
Lucent Search's research found that 88% of AI leaders find establishing effective AI governance challenging.
In some organisations, extending an established data-governance structure into AI will make sense. In others, AI governance now involves technology, legal, risk, operations and business leadership to such an extent that it cannot sensibly sit principally inside the data function.
Governance needs clear decision-making, appropriate authority and effective coordination across the functions involved.
When AI becomes a broader enterprise leadership question
AI does not only change how organisations analyse data. At scale, it can affect operating processes, customer journeys, workforce decisions, technology investment and risk across the organisation. As its reach widens, a separate leadership role can become easier to justify.
A CAIO may need to influence decisions across technology, data, operations, risk and the wider business. In that environment, success depends on much more than access to models and data. It depends on whether the executive can make different parts of the organisation work together.
Lucent Search's research found that 87% of AI and data leaders experience challenges with cross-departmental collaboration. Respondents described friction across IT, data and business functions, including competing priorities, fragmented data agendas and uncertainty over ownership.
93% cite strong demand for communication and influence skills.
Where the data leader is also expected to influence investment, priorities and change across the wider organisation, the demands of the role become quite different. The executive may bring deep expertise in data, analytics or AI, but enterprise execution increasingly depends on influencing investment decisions, resolving competing priorities and getting other parts of the organisation to change how they operate.
An existing CDAO who already operates effectively at that level may be well placed to lead AI. Where the data role is still predominantly functional, adding responsibility for enterprise-wide AI can represent a much greater change than the organisation initially assumes.
Even with the right person in place, the role may simply be too broad to absorb the additional responsibility.
Most organisations are dealing with overlap, not an empty space
In most organisations, responsibility for data and AI is already spread across several executives. The CIO may own enterprise platforms, integration and technology investment; the CTO may own architecture and engineering; and the CDO or CDAO may own data, analytics and some AI capability. Business leaders may sponsor important use cases and remain accountable for the operational or commercial results.
Lucent's research reflects how dependent AI delivery has become on coordination across IT, data and business functions. The difficulty often lies in the decisions between them, where ownership can become unclear and execution slows.
A CAIO can bring clearer ownership, but only if the responsibilities around the role are equally clear.
If a new AI executive arrives while the CIO still owns technology, the CDAO owns the data and analytics capability, finance controls investment, and the business owns adoption, there needs to be a clear reason for the additional role. What can that executive actually decide? Which responsibilities move with the appointment? Where do they have authority rather than simply another voice in the discussion?
If AI stays with the CDAO, that executive still needs the capacity and influence to carry the wider job.
Those dependencies remain whichever leadership structure the organisation chooses.
The role needs enough space to be done well
This is where I would spend most time before changing the leadership structure.
The first issue is the existing job. A data leader may already be responsible for a significant programme across governance, platforms, analytics, regulatory requirements and organisational change. Adding enterprise AI may create a role that is logically connected but practically too broad.
The strength of the team beneath the executive also changes what is realistic. Lucent's research found considerable variation in the size and design of AI and data functions, with nearly 40% of leaders managing distributed, multi-region teams and responsibilities ranging from relatively small specialist groups to substantial enterprise functions.
The same CDAO title can therefore describe very different leadership jobs.
A CDAO with experienced leaders running platforms, governance, analytics and AI has much more capacity to take on a broader remit. Where much of that responsibility still sits personally with the executive, adding enterprise AI can make the role unrealistic.
The role also needs enough authority to deliver what is expected of it. Being accountable for AI strategy is very different from being able to secure investment, influence technology decisions, resolve governance questions or challenge business functions that are not adopting the capability. If the executive cannot make or materially influence the decisions on which delivery depends, expanding their responsibilities will change very little.
Business ownership still needs to remain clear. Neither a CDAO nor a CAIO should become the place where the organisation deposits accountability for every outcome associated with AI. If AI changes a customer process, operating model or commercial decision, the relevant business leader still needs to own the resulting performance.
A separate CAIO should solve a real leadership problem
The strongest argument for creating a CAIO is that the organisation has identified a significant piece of enterprise responsibility that is not being led effectively within the current structure.
A separate CAIO may make sense where AI now spans several businesses and needs coordinated investment and governance, or where the existing data leader is already carrying a substantial programme. In other organisations, the gap may be a leader focused specifically on adoption and value across functions rather than the underlying data capability.
The strategic importance of AI does not, by itself, justify another executive role.
Broadening an existing CDO or CDAO role needs to involve more than adding AI to the job description. Responsibilities, authority, resources and expected outcomes may all need to change.
Sometimes the right answer is to strengthen existing leadership
Organisations can spend a great deal of time debating which new title they need when the better answer is to make an existing role work more effectively.
A CDO may already have the experience to lead analytics and AI but need a stronger team underneath them. A CDAO may already be carrying the AI agenda but need clearer authority across technology or the business. A CIO and CDAO may already have a workable division of responsibility that simply needs to be made more explicit.
Adding AI to an already substantial data role therefore needs careful thought. If the responsibilities expand materially, what changes in the team, authority and support around the executive?
Where the existing leadership does not have the capability required for the next phase, a new appointment may be necessary. Elsewhere, strengthening the team, clarifying responsibilities or giving the existing leader greater authority may be enough.
So, CDO, CDAO or CAIO?
CDO, CDAO and CAIO describe very different jobs from one organisation to another.
If the immediate challenge is improving ownership, quality and confidence in enterprise data, the senior data leader may still need to concentrate heavily on those foundations. If the organisation already has strong data leadership alongside analytics, data products and business adoption, AI may sit naturally within that responsibility as well.
A separate CAIO becomes more credible when AI has developed into a broader enterprise agenda that cannot be led effectively from within the existing data role — either because its reach across the organisation is fundamentally different, because the existing role is already too substantial, or because AI requires a different centre of executive attention.
The practical decision is whether one executive can lead data, analytics and AI effectively without the role becoming too broad or too dependent on decisions made elsewhere.
AI clearly deserves executive attention. Whether that requires another executive depends on the leadership already in place, how much the existing data role can reasonably absorb and what would actually improve by separating the responsibilities.
Related analysis and executive search
For organisations defining or appointing senior data and analytics leadership.
For organisations considering distinct enterprise AI leadership.
Do you need a Chief AI Officer — or should AI sit with the CIO, CTO or CDAO?
A broader look at where enterprise AI accountability should sit.
We’ve invested heavily in AI but aren’t seeing enough business value. What kind of leadership do we actually need?
A diagnostic for understanding whether the constraint is leadership, ownership, technology, data or execution.
Deciding how data and AI leadership should work?
Lucent Search works with boards and executive teams to clarify what existing leaders already own, where responsibilities need to change and whether a new appointment is actually required.
Where a search is the right answer, Lucent conducts the retained executive search.
Comments