At Gartner IT Symposium/Xpo 2024, analysts discussed four new challenges facing chief information officers (CIOs) in creating secure and scalable value with AI.

A survey of 451 senior technology leaders found that 57 percent of CIOs are responsible for their company’s AI strategy. A Gartner analyst highlighted that CIOs are often faced with high expectations due to constant technological innovation, while the reality of actually benefiting from AI remains complex.

“However, CIOs can set the pace in the race for AI results,” said VP Analyst Hung LeHong. “If you have modest AI ambitions in a sector that is not yet being transformed by AI, you can afford to take a more moderate pace. This is a steady pace of AI.”

“For companies with greater AI ambitions or in an industry that is being reinvented by AI, the pace will be faster. This is an AI accelerated pace. Whether you move at a steady or accelerated AI pace, you must deliver value and results.”

Below we explain the four AI challenges in more detail.

1. The business benefits of AI are not always clear

According to a Gartner survey, employees saved an average of 3.6 hours per week by using generative AI. But not everyone’s productivity is increased equally by AI. This increase varies greatly and depends on various factors. These factors are personal interest, level of acceptance, level of experience and the complexity of the position. This poses a challenge for CIOs responsible for delivering AI solutions.

Companies expect more benefits than just improved productivity. Improvements at operational and business levels also create added value. For example, automating important business processes or redistributing roles to work with chatbots.

It is important to manage the benefits of AI as a portfolio and determine the use of AI based on the specific areas where it can add value. Think about process improvements, automation or new revenue streams.

2. Cost management is crucial

The costs of AI can quickly spiral out of control. Especially when companies don’t properly understand the pricing structures and scaling of AI solutions. Gartner warns that CIOs could overestimate their AI spending by 500 to 1,000 percent if they don’t make the right assessment.

More than 90 percent of CIOs said cost management is a major challenge. CIOs must therefore understand cost components and pricing models and negotiate effectively with suppliers. LeHong emphasizes that conducting proofs of concepts should not only test the technology but also the cost implications. With a proof of concept you demonstrate the feasibility of a project.

3. Decentralization of AI and data

As AI and data become ubiquitous in the enterprise, they are no longer managed directly by the IT department. According to Gartner, on average only 35 percent of AI capabilities are now developed by IT teams. This requires new ways to securely manage data and AI assets.

Gartner therefore introduces the concept Tech sandwich. LeHong explains: “At the end of the sandwich is all the data and AI from IT, usually centralized. Added to this is the data and AI that come from everywhere, mostly decentralized. And in the middle are the TRiSM (Trust, Risk and Security Management) technologies that make everything secure. That’s what you need to create to accommodate AI and data coming from everywhere.”

This approach helps address the complexity of AI and data while leveraging new opportunities.

4. How does AI impact employee performance and well-being?

AI not only affects business processes, but also employees. Some employees feel threatened or upset by the rise of AI. This can lead to undesirable behavior such as jealousy or over-reliance on AI tools.

Only 20 percent of CIOs said their company is taking care to limit the negative impact of AI on employee well-being. Gartner recommends that organizations implement change management and clearly define who is responsible for managing behavioral outcomes.

These insights show that CIOs need a clear strategy to implement AI safely, cost-effectively and with human well-being in mind. This requires a structured approach that focuses on the balance between technological and organizational aspects.