By Zak Armstrong, Principal Presales Consultant, Experian
Artificial intelligence (AI) is moving from experimentation to operational reality, but organisations are still struggling to build the trusted data foundations needed to support it. Experian’s Connected Intelligence research found that 62% of respondents agree that data quality and governance are among the reasons AI deployments fail, while 54% report increased demand for high-quality, structured data as AI adoption increases. For data governance leaders, the challenge is no longer whether to adopt AI, but whether their data is ready to support it.
There is no shortage of enthusiasm for AI. Across organisations, teams are exploring new use cases, activating new technologies and looking for ways to improve decisions, automate processes and work more efficiently. However, investment in AI is often moving faster than the work required to understand and govern the data behind it.
I recently explored this challenge with Belinda Toniolo during a fireside chat at the AI & Data Governance Summit in Melbourne. Belinda joined the discussion in a personal capacity and brought a practical perspective shaped by extensive experience across data governance, quality and enablement.
Our discussion reinforced an important point: AI readiness is not only about having the right technology. It is about knowing whether the data supporting an AI initiative is understood, trustworthy and fit for its intended purpose.
This reflects what Experian is seeing more broadly across the market. Our Connected Intelligence research found that 68% of financial institutions remain in the early or emerging stages of AI adoption, despite significant investment and interest in AI-driven decisioning. This suggests many organisations are still building the governance, quality and trust frameworks required to scale AI with confidence.
AI readiness means different things to different teams
Ask several teams what it means to be AI-ready and you are likely to receive several different answers. A security team may focus on access and protection. A legal team may consider obligations and risk. A technology team may look at infrastructure, models and platforms. From a data governance perspective, the question is more fundamental. Is the data ready for the way the organisation intends to use it?
An AI system can only work with the information it can access. Organisations therefore need to understand the data being provided, its quality, its meaning and any limitations that could affect an outcome.
Experian’s Connected Intelligence research reinforces the scale of the challenge. Data quality, governance, integration and access continue to be barriers to scaling AI, while demand for structured, high-quality data is increasing as organisations expand AI use cases.
The implication is clear: organisations can modernise tools and models, but without trusted data foundations, scaling AI remains difficult.
This principle is not unique to AI. Data has always needed to be fit for purpose, whether it is being used by a person, a reporting process or an automated system. AI may allow data to be used across some processes at greater speed and scale. This may make existing data quality gaps more visible and increase their effect on downstream outcomes.
Start with the outcome, not the technology
AI conversations can quickly move towards what might be possible. That ambition is useful, but it needs to be connected to a clear business outcome. Before progressing an initiative, organisations should ask:
- What problem are we trying to solve?
- What outcome do we want to achieve?
- What data will the initiative require?
- Do we capture that data today?
- Is the available data suitable for this purpose?
Sometimes, the data an organisation wants to use does not exist. It may not be captured, it may be incomplete, or it may not contain the detail required to support the proposed outcome.
AI may help organisations identify patterns and accelerate some processes, but its usefulness will still depend on the availability and suitability of the underlying data. Organisations are recognising this challenge. Experian’s Connected Intelligence research found that 75% of Australian organisations expect AI, data and software investments to deliver faster or real-time decision-making, yet 68% also agree that data quality and governance are among the reasons AI deployments fail. The message is clear that organisations may be investing in AI to move faster, but success depends on having trusted, well-understood data foundations in place. Governance teams can add considerable value at this stage. By connecting business ambition with an informed view of the organisation’s data, governance can help teams design more achievable use cases and identify the work required before implementation.
You don’t need to govern everything at once
Most organisations hold more data than any single team could govern in detail. Attempting to address the entire data environment before progressing an AI initiative is unlikely to be practical. The better starting point is to focus on the data that matters most. Identify the business processes and outcomes most likely to benefit from AI. Then determine which data supports those processes and focus governance activity there.
For critical data, organisations should be able to explain:
- What the data means
- Where it comes from
- How it moves through the organisation
- Which processes create and consume it
- Who owns it
- What level of quality is required
- Whether its current quality is suitable for the proposed use
This targeted approach allows smaller governance teams to focus limited time and resources where they can have the greatest impact. The need for prioritisation is becoming more apparent as AI adoption grows. Experian’s Connected Intelligence research found that 62% of Australian organisations report increasing demand for high-quality, structured data as AI use expands. For governance teams, this reinforces the importance of focusing on the datasets that are most critical to business outcomes rather than attempting to govern everyone at once.
AI readiness does not require every dataset to be perfect. It requires the organisation to understand the data that matters to a specific outcome and make an informed decision about whether it can be trusted.
Governance must become a business enablement function
At Experian, we increasingly see governance as an enabler of trusted AI rather than simply a control function. While policies, controls and accountability remain essential, organisations that are making the most of progress are often those that connect governance initiatives directly to business outcomes. Rules along won’t help organisations move confidently into AI. Governance needs to help the business understand data, manage risk and accelerate responsible innovation.
Governance teams increasingly need to operate as business partners. That means understanding organisational strategy, recognising where AI is likely to be applied and connecting teams that may otherwise work separately.
One business area may hold valuable data that another team does not know exists. A technical team may know that a dataset has quality issues but lack the context to understand why those issues matter. Senior leaders may have an ambitious idea without visibility of the data required to deliver it.
By bringing together business context, data knowledge and clear accountability, governance teams can help the organisation make better decisions and progress more quickly. This shifts the role of governance from controlling access to enabling appropriate and informed use.
Confidence starts with an honest assessment
Organisations do not need to claim that their data is perfect before exploring AI. They do need to understand its condition and communicate that honestly.
A practical readiness discussion should cover five questions:
- What business outcome are we trying to achieve?
- What data will the initiative rely on?
- Do we capture and understand that data today?
- Is its quality suitable for the intended purpose?
- Can we explain its ownership, meaning, lineage and current use?
If an organisation cannot yet answer these questions, that does not mean the AI initiative must stop. It means there is clear work to prioritise. AI readiness is not a destination reached through one framework, platform or policy. It is the ability to make informed decisions about how data can be used, what risks need to be addressed and where improvement is required.
Trusted AI starts with trusted data. More importantly, it starts with understanding what trusted means for the outcome the organisation is trying to achieve. Trust and transparency are emerging as critical success factors for AI adoption. Experian’s Connected Intelligence research found that 86% of organisations consider transparency in analytics and insights highly valuable for improving decisions. Organisations that can clearly explain their data, governance processes and AI decision-making will be better positioned to scale AI responsibly and with confidence.
Next step
If you would like to discuss your organisation’s data quality, governance or AI-readiness priorities, speak with your Experian account manager or complete the client enquiry form.
Frequently asked questions
What does AI readiness mean for data?
AI readiness means understanding whether the data required for a specific AI initiative exists, is understood, is of suitable quality and can be used responsibly for the intended business outcome.
Why is data quality important for AI?
AI systems rely on the data available to them. Incomplete, inconsistent or poorly understood data can reduce confidence in outputs and make existing data issues more visible when processes operate at greater speed or scale.
Does an organisation need to govern all its data before using AI?
No. A practical approach is to begin with the intended business outcome, identify the critical data supporting it and focus governance activity on that data first.
What data governance questions shouls organisations ask before scaling AI?
Organisations should ask what outcome they want to achieve, what data the initiative will rely on, whether that data is captured and understood, whether its quality is fit for purpose, and whether its ownership, meaning, lineage and current use can be explained.
How can data governance support responsble AI adoption?
Data governance can connect business goals with data quality, ownership, lineage and risk. This helps teams identify gaps early, prioritise improvement work and make more informed decisions about whether an AI use case is ready to progress.
What is trusted data?
Trusted data is data that is sufficiently understood, accurate, complete, governed and fit for a defined purpose. The level of trust required depends on the business outcome and the risks associated with the proposed use.