Data and IT leaders are under mounting pressure to turn artificial intelligence experiments into tangible business results. Despite the hype, and growing fears of an AI bubble, analysts argue that the real challenge is not technological but organisational. To succeed, organisations must move beyond simple return-on-investment calculations and focus on governance, cost transparency, and workforce readiness.
Speaking at a recent data and analytics summit in Sydney, analyst Jorg Heizenberg described AI as a shift as profound as the arrival of the internet. That comparison underscores the scale of change confronting businesses. AI is not just another tool to be bolted onto existing processes; it changes how decisions are made, how work is performed, and how value is created.
Georgia O'Callaghan, a director-analyst at the same firm, highlighted the urgency. Nearly three in five organisations had an AI service in production in 2025, and four in five are now increasing their investment. But, she warned, 'you can't just continue to increase your investments in AI without getting clarity on the goals and ambition of your organisation.'
Set ambitions and risk appetite
Heizenberg said data and analytics professionals should redefine their AI ambitions with stakeholders, particularly regarding tolerance for disruption. Those with low tolerance can assess risk and follow the safest course. Those with greater appetite can take an opportunistic approach. High-tolerance organisations might dare to be pioneers, accepting the biggest risks.
The key is that ambition must be explicit. Without a shared understanding of whether the organisation is trying to optimise existing operations, enter new markets, or transform its business model, investment decisions become incoherent and results are hard to measure.
AI costs are unpredictable and often hidden
One of the first questions stakeholders ask is 'What is this going to cost?' But that is difficult to answer. AI costs are highly unpredictable and often hidden. Vendors use pricing based on metrics such as GPU hours and token consumption, which are hard to forecast.
Gartner's research shows a troubling disconnect: six out of ten IT leaders worry about AI agents running up unexpected costs, but only two out of ten data and AI leaders are concerned that unpredictable pricing might limit the value they get from AI. Heizenberg called this a wake-up call.
Less than half of organisations manage and optimise their AI-related spending, O'Callaghan noted. She advised tracking expenses from the outset, especially during prototyping, and adopting cost-driven design. Teams should examine the cost implications of using different large language models or small language models to power an AI agent.
However, when communicating with stakeholders, the focus should remain on value rather than just cost. Value is more than money. Heizenberg gave the example of North Yorkshire Council, which created a digital citizen named Dotty and mapped her journey through public services to make the impact of data relatable. If two digits in a home address are transposed, a tradesperson might be sent to the wrong house to install a handrail for an elderly person. The wasted journey has a direct financial cost, but the ripple effects could be far worse if the lack of a handrail leads to a serious fall.
Foundational investments pay off
Whatever an organisation's ambition, foundational investments are essential. A 2025 survey on modern data realisation found that respondents most satisfied with their AI outcomes spent 30% more on foundational activities such as data management, governance and talent than those who were unsatisfied. That suggests shortcuts in data quality or governance eventually show up as disappointing AI results.
Yet many feel pushed into AI before they are ready. In Gartner surveys, 59% of IT leaders said they were being pushed into adopting generative AI tools before ready, and 61% felt pressure from senior leaders or stakeholders to move forward. The antidote is not to slow down innovation but to build the data and governance foundations that make innovation safe.
Governance as a value accelerator
One of the biggest concerns is whether an organisation's data is secure and well-governed enough to be opened up to further AI applications, including autonomous agents. 'We need to prevent the exposure of the wrong data to the wrong people, applications or LLMs with AI governance, and avoid inaccuracies, misunderstandings and hallucinations with a well-designed context layer,' said O'Callaghan. This ensures data is AI-ready, trusted and aligned to the use case.
Governance should be repositioned as a business value accelerator rather than a compliance function. Analysts suggested three steps. First, connect existing governance groups such as risk, data and cyber security into a unified AI governance team. Gartner predicts that organisations doing this will experience a 10% greater business impact than those that do not.
Second, rationalise governance. The unified team should review and consolidate policies into a clear, consistent framework that reflects the organisation's risk tolerance and cultural values around responsible AI use. This reduces confusion and ensures that rules are aligned with strategy.
Third, embed governance across both the culture and the technology of the business. That requires shifting the organisational mindset from compliance to one where everyone understands how to use data responsibly and ethically. Technologically, leaders should adopt policy-as-code so rules are automatically enforced throughout the tech stack. Gartner predicts that by 2028, organisations using specialised governance tools will decrease the cost of regulatory compliance by up to 20%.
Context is critical
Even well-governed data can be misinterpreted. Context matters. If an employee asks how many active customers the business has, the answer depends on the definition of 'active'. Does it mean someone who made a recent purchase, holds an ongoing subscription, or recently visited the website? In the absence of context, an LLM can misunderstand the prompt and rapidly amplify the error.
Heizenberg argued that it is time to build an integrated context realisation layer – a layer that connects every piece of information so everyone and everything, people and agents alike, can see the bigger picture and make more informed decisions. Semantic layers are becoming common, but they are no longer sufficient on their own. Organisations are experimenting with ontologies, knowledge graphs and other methods to attach deeper meaning to data. Combining these approaches yields far more accurate results.
Investing in people
Another challenge is that technology is evolving faster than the workforce can adopt it. 'If you're investing in AI without investing in your people, you are throwing money away,' warned O'Callaghan. The change management and training effort for AI tools takes nearly twice as long as implementing the AI solution itself. That means planning for longer timelines and higher costs than for any other technology implementation.
A 'mindset, skillset, toolset' approach is effective. IT leaders must ask what mindset obstacles exist and how to overcome them; what skills gaps are present and how to remedy them. Only after addressing mindset and skillset should leaders consider tooling changes.
There is also the ongoing concern about AI-driven job losses. Gartner found that 34% of CIOs expect to reduce their workforce over the next three years. Conversely, only 4% of chief data officers have decreased their team size in the past year, while 44% have expanded their teams. O'Callaghan noted that some organisations may be using AI as a convenient excuse for layoffs that would have occurred regardless.
The value of human skills will still sit at the core of delivery teams, but these teams will combine human expertise with AI agents to make more productive, AI-powered fusion teams. The future is not about replacing people but about augmenting them with intelligent tools that can handle routine tasks and surface insights.
Source:ComputerWeekly.com News