Qlik Q&A: Why CAIOs Are Essential in Modern Finance Firms

With AI shifting from an experimental tool to an essential part of core operations, C-suites are increasingly recruiting AI-specific execs in line with the need for executive ownership of AI.

Because AI is now a central part of modern business strategies, Chief AI Officers are becoming essential for managing AI risk, ethics and investment across financial services and insurance.

An IBM report found 76% of those it surveyed now have a Chief AI Officer, up from 26% just a year ago

HSBC’s appointment of its first Chief AI Officer is one of the more recent signs that banks are moving to formalise how AI is governed, developed and deployed. 

As regulators and policymakers intensify scrutiny of AI-related risk, financial institutions are under growing pressure to show clear accountability, strong data controls and transparent decision-making frameworks.

Against that backdrop, Martin Tombs, Field CTO EMEA at Qlik, argues that the success of AI in finance will depend less on experimentation and more on disciplined foundations: governed data pipelines, explainable outputs and alignment with risk and compliance structures. 

In this Q&A with FinTech Magazine, he explains why fragmented data remains one of the biggest obstacles to trustworthy AI, what regulators want to see from firms using AI and the practical steps banks need to take to embed AI responsibly at scale.

What does HSBC’s appointment of a Chief AI Officer reveal about how banks are formalising AI governance?

HSBC’s appointment of a Chief AI Officer points to a broader move within banking.

AI is moving out of isolated pilots and into a more structured, organisation-wide model, with clearer ownership over how it is governed, developed and deployed. 

Financial services, like many sectors, have traditionally taken a cautious approach to AI, reflecting the highly regulated environment in which they operate.

While AI has been on the agenda for some time, recent developments point to a more deliberate shift toward ensuring it is implemented responsibly and in a way that protects both the stability of the industry and the security of its customers.

This shift is becoming increasingly visible across the sector. 

More firms have recently joined the Financial Conduct Authority’s initiative to test AI in real-world conditions under strict controls, while the Bank of England has outlined plans to assess potential risks to financial stability through scenario analysis and simulations. 

These steps follow earlier criticism from the Treasury Committee, which argued that regulators had been too cautious and had adopted a wait-and-see approach for too long. 

What we are seeing now is a more forward-looking approach to AI in financial services, and one where governance and regulation are firmly at the centre.

How can financial institutions ensure their data foundations support reliable, auditable AI outcomes?

Financial institutions can only deliver reliable, auditable AI outcomes if their data foundations are strong enough to withstand scrutiny from the outset, rather than being retrofitted after AI models are built.

In practice, this starts with tackling data fragmentation. Financial data is often spread across different systems, creating silos that make it difficult to trace how inputs flow into AI outputs. 

Without clear data lineage, it becomes significantly harder to explain, validate or defend AI-driven decisions under regulatory scrutiny.

This issue is magnified by the fact that 39% of organisations still operate with incomplete or limited internal data. 

That naturally limits the quality of insights and makes it harder to build trust in AI outputs.

Addressing this means bringing data together in a more unified and governed way. Data quality needs to be actively managed and context needs to be preserved as data flows between systems so it remains meaningful and usable.

Ultimately, AI outcomes are only as reliable and defensible as the data pipelines that support them. When those pipelines are well-structured, transparent and governed, institutions are in a far stronger position to produce AI outcomes that are accurate, explainable and able to stand up to regulatory review.

What are the biggest barriers to breaking down siloed data environments in finance?

One of the main challenges for organisations across all sectors is deciding where to focus their limited time and resources. 

Many often do not have the bandwidth to connect everything at once, and trying to do so can actually slow progress down rather than improve it. Because of that, teams are often forced to prioritise certain data sources over others, which means some useful information inevitably stays disconnected, simply because it is not seen as urgent enough to tackle. 

In financial services companies, this can show up as fragmented views across risk, compliance and customer data. For example, transaction data may sit separately from fraud or AML systems, which limits the organisation’s ability to build a complete, real-time picture.

There is also the difficulty of identifying what data is actually worth integrating. With so much information being generated across the business, it is not always clear which datasets will drive better decisions and which will add complexity without adding much value. 

Addressing this comes down to assigning clear ownership for key datasets, so there is accountability for how they are managed and integrated and making sure teams are aligned on which data matters most.

That way, integration efforts stay focused on areas that will actually improve decision-making, rather than trying to connect everything at once.

Who should own accountability for AI-driven decisions in a regulated organisation?

Responsibility for outcomes cannot sit in isolation with any single group.

In regulated organisations, technology teams, data teams, AI specialists and business stakeholders all share an obligation to understand the importance of data and the consequences it has on decision-making.

Technology and data functions play a critical role in ensuring models are built on trusted, well-governed data and operate consistently, while other business users remain accountable for how those insights are applied and the outcomes they ultimately drive.

In that sense, accountability is distributed, but it only works when there is shared responsibility for data quality, governance and the impact of AI-driven decisions across the entire organisation.

What does responsible, scalable AI adoption look like for financial services over the next few years?

Over the next few years, responsible AI in financial services will really be about moving from pilots to scale but doing it in a controlled and governed way.

That starts with data and without trusted, unified data foundations, organisations will struggle to scale AI without creating risk and inconsistency.

We will also see governance becoming much more embedded into the design of AI systems, rather than something that sits around them. 

And finally, it is not about full automation, but about finding a balance. 

AI should support faster, better decisions, but still with clear human oversight where it matters.