Everyone’s betting on AI, but few are integrating it well

FOR business leaders under pressure to deliver results amid rising costs and economic headwinds, the constant stream of artificial intelligence (AI) innovations can feel like the obvious solution. Every week brings the “next big thing”, promising faster and smarter results.

Yet, this bullish excitement can backfire quickly when organisations rush to adopt without considering integration. International Data Corporation’s research shows that one in five AI projects in the Asia-Pacific fails, often because tools cannot be connected to existing systems and data. Many more remain stuck in pilots, unable to scale. The result is wasted investment, fragmented workflows and frustration.

Businesses today operate in complex environments built on legacy systems, packaged software, and now, layers of AI tools. Left unmanaged, this leads to “AI sprawl” – the unchecked proliferation of tools across functions. It is a familiar pattern. Much like the shadow IT era, when employees adopted unapproved apps that created silos and complexity, today’s AI tools risk creating the same governance and compliance blind spots, only at far greater scale with AI.