Despite increasing AI spend, firms falter in strategic execution, report finds

A new report commissioned by HPE Aruba Networking has revealed that despite having strong faith in their AI strategies, many organisations display significant gaps when it comes to execution. The report "Architect an AI Advantage" suggests that businesses are failing to grasp the complexities of an end-to-end AI life cycle. Less than half of IT leaders fully understand the various demands of AI workloads, while scanty percentages have set up data governance models or can operate real-time data pushes/pulls. Moreover, the study reveals that 22% of IT leaders fail to involve legal teams in their AI strategy discussions.

According to Sylvia Hooks, VP of HPE Aruba Networking, the findings highlight a paradox in the evolving landscape of AI adoption. "AI adoption is picking up pace, with nearly all IT leaders planning to increase their AI spend over the next 12 months," she said. However, despite this appetite for AI, Hooks notes "very real blind spots that could see progress stagnate if a more holistic approach is not followed." Such misalignments could prevent organisations from fully leveraging areas of expertise and making efficient decisions.

The study showed a worryingly low level of data maturity, a critical component for AI success, within organisations. A mere 7% of organisations can operate real-time data pushes/pulls, while only 26% have implemented data governance models and can conduct advanced analytics. Surprisingly, less than 60% of respondents were confident in their organisation's ability to handle key stages of data preparation for AI models, such as accessing, storing, processing, and recovering data. Such data-related challenges could slow down the AI model creation process and increase the likelihood of inaccurate insights and poor ROI.

Confidence levels regarding networking requirements and system capacity across the AI lifecycle appear high among IT leaders, with 93% believing their network infrastructures can support AI traffic. However, less than half of these leaders claim to fully understand these requirements, flagging potential issues around accurate provisioning.

The report also pointed out a lack of cohesion in AI strategy across different business areas, with 28% of IT leaders describing their approach as "fragmented." The amalgamation of compliance and ethics into AI strategy is worryingly absent, a discrepancy that risks a backlash from increasing scrutiny on these aspects from both consumers and regulatory bodies. Research indicated that almost a quarter of organisations (22%) do not involve their legal teams in AI strategy discussions at all.

The report highlights that while the haste to adopt AI is understandable, given its wide-ranging benefits, ill-considered strategic planning can expose businesses to considerable risks. These can stem from a breach of unprotected proprietary data, poorly constructed AI models, or even unnecessary increases in data centres' carbon emissions. Dr Eng Lim Goh, SVP for Data & AI at HPE, stressed that despite AI's "enormous potential", businesses must carefully consider the risks and ensure they fully understand the complexities across the AI lifecycle to avoid jeopardising their ROI.