FEATURE where AI is designed to augment employees rather than replace them.
“ Over the past year, organisations have moved beyond experimentation in areas where AI supports people rather than replaces them. We’ re seeing real adoption in customer service, employee support, IT operations, software development, and knowledge management – delivering measurable productivity gains.
“ Where expectations outpace reality is clearer: autonomous decision-making and end-to-end process ownership. LLMs excel at generating and analysing information, but lack the contextual understanding required for highconsequence decisions.
“ The organisations seeing greatest success embed AI into workflows where it handles repetitive tasks while employees focus on judgement and decision-making. However, I’ ve sat in multiple conversations with leaders asking the same thing:‘ why aren’ t we seeing the value we expected?’ Many assume the technology is the problem.
“ I believe it lies elsewhere. The areas where we are seeing adoption succeed share one thing: they’ re contained. They don’ t require seamless data flows across silos or coordinated decisions across disconnected systems. The moment organisations try to scale AI beyond these boundaries, the structure breaks.
“ This isn’ t a technology problem. It’ s an enterprise architecture problem. AI isn’ t the transformation. The enterprise is.”
Joe Dunleavy, Regional CTO and Global Head of Dava. X AI Group at Endava, believes the popularity of AI-assisted software development illustrates both the opportunities and limitations facing enterprises today.
“ We’ ve moved well past experimentation in some areas, and vibe coding is the clearest example. www. intelligentcio. com
INTELLIGENT CIO MIDDLE EAST
25