HCLTech warns 43% of major AI initiatives may fail

HCLTech has released research suggesting 43% of major enterprise AI initiatives are expected to fail. The study is based on responses from 467 senior leaders at large companies across 10 countries.

The report highlights a widening gap between rapid AI adoption and organisations' ability to turn that investment into sustained business value. The challenge is not a lack of experimentation or access to tools, but difficulty delivering consistent results across large organisations.

In Australia, the findings identify legacy integration, data modernisation and governance gaps as key barriers to scaling AI. They also show that 76% of organisations believe both legacy and already-modernised applications need major overhauls to properly support AI.

This is adding pressure on technology and business leaders as expectations for returns tighten. Nearly half of enterprise leaders surveyed expect measurable value from AI investments within 18 months, leaving little room for delay or missteps.

The research focuses on companies with more than USD $1 billion in annual revenue and covers AI use in IT operations, software engineering and business functions. As AI moves deeper into core operations, the report argues, failures become more visible and carry greater strategic risk.

Australian pressures

Australian respondents also reported strong competitive pressure to move faster. The study found that 89% of leaders believe competitors will be using autonomous AI systems for mission-critical work within the next 12 months.

At the same time, organisations are struggling to align AI programmes with broader business strategy and build convincing business cases for large-scale investment. Initiatives are more likely to stall when technology teams and business leaders are not aligned on objectives and execution.

The findings also point to the state of enterprise applications and operating models. For Chief Information Officers and other technology leaders, AI deployment is exposing weaknesses in systems and data environments that were not built for autonomous or continuously learning tools.

For senior executives outside technology functions, the issue looks different. Aggressive AI spending without organisational alignment can create strategic risk, particularly when projects move quickly into critical parts of the business.

Change management

The study identifies change management as one of the most important factors in determining whether AI projects succeed. Yet it remains one of the least funded areas in many enterprise AI programmes.

Many organisations are deploying AI into workflows without preparing the staff expected to use it alongside their day-to-day work. The report describes that gap as a primary execution risk as companies try to scale adoption under tighter timelines.

Governance also features strongly in the Australian results. Some 79% of organisations said governance and responsible AI considerations significantly influence deployment decisions, reflecting concerns about accountability and oversight as the technology is applied more widely.

The report also points to growing interest in agentic and physical AI use cases, including applications in manufacturing, engineering and operations, rather than only digital workflows. These uses remain at an early stage, HCLTech says, but they raise further questions about reliability and control.

That shift broadens the debate beyond software tools and productivity gains. The next challenge for large companies, the report suggests, will involve not just technical integration, but decision-making structures, risk management and workforce preparation.

Vijay Guntur, Chief Technology Officer and Head of Ecosystems at HCLTech, commented on the findings.

"AI has moved from being a technology initiative to becoming an enterprise operating reality. What leaders are grappling with now is not whether AI can deliver value, but how organisations adapt their structures, decision rights and risk tolerance to keep pace with it. The pressure to move fast is real, but without the right investment in people, in helping them understand, trust and work effectively alongside AI, speed can just as easily amplify failure as success," Guntur said.