AI adoption: How CIOs can balance innovation with financial discipline

The true cost of AI adoption

Demand and excitement for generative AI cannot unravel the financial acumen of technology executives. 

The success of artificial intelligence (AI) will not be defined by the implementation of the technology, but rather by the business value it delivers for the organisation.

After all, despite the many proclaimed productivity benefits of the latest wave of generative AI and now agentic applications, these technologies must drive bottom-line results. If they don’t, artificial intelligence could become a significant financial burden to the enterprise. 

Fortunately, this is not the first time CIOs, CTOs and CDOs have faced a wave of technology optimism with questionable financial value.

In the early 2000s, technology executives found themselves navigating the early internet boom—when enterprises rushed online without clear financial oversight. Years later, many organisations were caught in the rush to SaaS, cloud computing and the early days of Hadoop and Big Data.

During these times, some technology teams, leaders and company CEOs lost sight of the business goals, and IT largesse became accepted in many quarters. Today, however, in the face of economic hardship and rising computing and AI costs, CIOs must ensure AI investments align with business goals, avoiding unchecked IT spending which could erode financial stability.

The cost of AI: What enterprises are facing

The average cost of enterprise computing is expected to increase by 89 percent, according to the Institute of Business Value, an IBM research organisation - and this is being driven in-part by the growing cost of resource-intensive AI. 

Approximately 70 percent of business executives cite AI as one of the causes of this cost increase, while the report also finds that many organisations have cancelled or postponed at least one generative AI project due to cost concerns. This approach is unsustainable, as it will be hard to prove where AI can, or cannot, make a difference. 

Technology analysts Gartner describe cost concerns as “the greatest near-term threat” to AI success. Its research finds that organisations spent, on average $2.3 million in 2023 on generative AI proof of concepts (POC). More recently, Goldman Sachs’ report into genAI has suggested that the spending on generative AI has ‘little to show so far’ in the way of benefits and returns.

These rising AI costs sit uncomfortably against a backdrop of trade wars, geopolitical conflict and the resulting supply chain challenges - all of which puts pressure on technology executives to ‘do less with more’.

Investing in AI: Driving cultural change at PZ Cussons

Despite these concerns, technology executives and their organisations do not want to miss opportunities to keep pace or even overtake the competition. Previous research from Cisco has found that 70 percent of CEOs are worried that they could lose out to competitors if they do not invest in AI. 

UK-based consumer goods firm PZ Cussons is an early adopter of Microsoft Copilot, the generative AI assistant technology. Like many organisations, the real productivity benefits of Copilot have not been realised just yet, but CIO Jawaz Illavia has a unique perspective on the adoption of these AI technologies - and what return on investment (ROI) could be.

“Copilot adoption was part of the bigger picture of doing things differently and creating a more technology-minded workforce,” Illavia told HotTopics. “We did technology ‘jam’ events focusing on the latest technology trends and to show people what the industry and what we are doing. 

“The buzz this created meant that we had employees – who were never interested in tech trends – all of a sudden downloading whitepapers to learn more; something they had never done before. 

“With Copilot, people say, ‘what is the RoI?’ Sometimes, the RoI is getting people up to speed, thinking differently, being more curious and exploring new ways of solving problems. Maybe this new tool doesn’t currently have a great RoI as yet, but they may apply the same principles to the next tool, so I used it as a catalyst for cultural change.” 

The CIO doesn’t apologise for using the hype created by the arrival of generative AI to trigger the cultural change.

“There are very few technologies out there that have been able to capture so much imagination; it was one of the best chances for technology leaders to engage at scale and get people excited. It has created pull rather than push. 

“With Blockchain and Metaverse technologies, nobody really understood them, but with generative AI, people understand it.  People will adopt if they understand technology, but if they don’t understand it, they will block and resist. 

“Also, people are time-poor; they need to grasp it quickly and with minimal investment of their time.” 

Illavia has invested in a cultural change, which is a key component of any technology transformation. Too often in the history of enterprise technology, organisations have invested in the tool, but not the change management required to get the return on investment. 

Robert Pick, CIO of insurance firm Tokio Marine North America, believes AI transforms how the workforce engages with technology, and therefore the conversation about change management.

“The way we are using systems will change, the inflection point for human interaction will evolve,” he said.

Measuring AI costs and ROI: Learning from cloud 

If cultural investments are, to a degree, intangible, CIOs will still need to work with their finance peers on tangible cost management as part of the adoption and investment into AI - yet this alignment is anything but straightforward.

According to a KPMG report, nearly 1 in 3 CIOs say their tech budgets are currently insufficient, but just 12 percent of CFOs think the same. At the same time, the cost of IT services is growing, exacerbated by economic challenges and the cost of resource-intensive technologies like AI and cloud computing.

CIOs have faced a similar challenge before with the adoption of cloud. Like AI, this iteration of technology has been largely easy to adopt. As a result, organisations have seen their cloud costs soar, particularly as companies worked with multiple providers across their estate. Those of us who were observing this technology in its early days will recall an initial promise for cloud to reduce costs, certainly when compared to on-premise data centres.

The truth, however, is like AI, cloud computing is not about cost reduction but enabling a different and more flexible way of working.  

Tokio Marine’s Pick sees similar adoption trends moving forwards.

“AI will be an incremental change, not a revolution.  Generative AI is revolutionary technology that will be applied in an evolutionary fashion.” 

Keeping AI costs under control

As these technologies evolve and impact the bottom line, CIOs have learnt to adopt methods which enable them to fully analyse and manage costs. 

Cloud and data consulting organisation esynergy is a specialist in Cloud FinOps, a model for improving cloud cost management; it finds that 59 percent of organisations are increasing their visibility and monitoring of cloud costs. The reason is to create the right culture towards cloud usage. 

This same approach will become vital to CIOs looking to manage the cost of AI adoption. As a technology built on the cloud, many of the pillars of Cloud FinOps will help CIOs drive cultural change and embrace agents and generative AI – while keeping a staunch eye on the budget. 

Key takeaways:

1. AI adoption must deliver business value – AI investments should be measured by their impact on revenue, efficiency, and competitive advantage, not just technological implementation.
2. AI costs are a growing concern – Enterprise computing costs are rising sharply due to AI, with many businesses already postponing projects due to financial constraints. CIOs must ensure AI cost control and be focused on delivering clear ROI.
3. Cultural transformation is key – AI success isn’t just about tools; it’s about mindset. CIOs should use AI to foster innovation, upskilling and engagement across the workforce.
4. AI’s ROI may not be immediate – The benefits of AI adoption (like Microsoft’s Copilot) might not be immediately financial but can drive long-term gains in productivity and adaptability.
5. Learn lessons from cloud’s adoption – Just as cloud uptake led to unforeseen costs, AI requires careful financial oversight, leveraging FinOps principles for budgeting and optimisation.
6. AI will be evolutionary, not revolutionary – Despite AI’s potential, most businesses will integrate it incrementally. CIOs should take a measured, strategic approach rather than rushing into large-scale deployments.