Peter Bristow, Change, Transformation & AI Leadership Search Partner, Executive Recruit

Fear is driving investment decisions

A CEO, gripped by the fear of being left behind, pounds the table and declares, ‘We need AI.’ When a subordinate tentatively asks what specific problem the technology should solve, the response is a dismissive, ‘Just add AI to everything.’ This anxiety-driven decision-making is currently costing businesses billions.

The pressure is immense, often emanating from board members who may not fully grasp the technology’s nuances but are insistent on visible AI adoption. The temptation to jump on the bandwagon, to mimic competitors and appease stakeholders, is powerful, but the evidence is clear: this approach is a fast track to wasted capital and unfulfilled promises.

The core problem is that leaders are treating artificial intelligence as a shiny, plug-and-play tool rather than what it truly demands: a profound, structural transformation of how their entire business operates. The data is sobering, a staggering 94% of businesses investing in generative AI at scale are seeing zero return on their investment.

Three pitfalls that doom AI projects

This hasty rush to adopt AI without a clear strategy results in several critical pitfalls that doom projects from the start. Perhaps the single most common and egregious mistake is the relentless pursuit of speed without a corresponding focus on quality or business value. Companies are enthusiastically using AI to accelerate workflows that they should be eliminating entirely.

Applying artificial intelligence to an inefficient or broken process does not create a transformed operation; it simply makes the broken process faster. This focus on superficial efficiency rather than fundamental process redesign is a hallmark of the bandwagon approach, where the goal is simply to implement AI rather than to solve a real business problem.

Compounding this problem is a severe misallocation of resources, where budgets are heavily lopsided. Many organisations dedicate over 90% of their AI budgets to acquiring sophisticated software and infrastructure, leaving less than 10% for the critical work of training the workers who will actually have to use these new tools.

This ‘funding tools over people’ mindset creates a situation where a company possesses powerful technology but lacks the human capability to wield it effectively. The transformation is doomed from the start because the workforce is unprepared, anxious, and untrained. This lack of focus on people leads directly to another major failure: the ‘pilot trap.’

Executives frequently launch disconnected proof-of-concept projects in single departments, often to generate press releases or satisfy board demands. However, because these projects lack an enterprise-wide vision and are not integrated into core workflows, they inevitably stall out. They create fragmented data silos and generate zero measurable impact on the bottom line.

Organisations that build AI capability through occasional training programmes see less than 18% of employees report genuine confidence in the AI systems they are asked to use. In stark contrast, in organisations where AI is woven into everyday work with role-specific examples and dedicated time for experimentation, that figure rises to more than 50%, highlighting that participation and ownership are what actually build trust and drive adoption.

A leadership failure, not a technology failure

The true cost of this bandwagon approach goes far beyond the billions wasted on failed pilots and unused software. It represents a profound leadership failure, a fundamental misunderstanding of what it takes to create lasting value from a transformative technology. The prevailing justification for incorporating AI is often that ‘everyone else is using it’.

This is not a strategy; it is a capitulation to peer pressure and hype. When leaders ask, ‘What if my competitor solves it first?’ instead of ‘What problem am I solving?’ they are abdicating their primary responsibility to the business.

This reactive stance leaves companies vulnerable to the influence of external consultants who may frame slow adoption as a ‘career risk’ for executives, bypass technical leaders with legitimate security concerns, and use cherry-picked data to create a bandwagon effect that drives massive, often unnecessary, deals. The result is a cycle of waste and disappointment that could be avoided with a more disciplined and value-first approach.

The value-first way forward

The most successful executives have escaped this cycle by adopting a measured, outcomes-focused framework. They look past the hype and focus heavily on business readiness, strict governance, and long-term process transformation. The optimal strategy begins with a fundamental shift in mindset: leaders must stop starting meetings by asking, ‘How can we use AI today?’ and instead ask, ‘What are our biggest operational bottlenecks, and can AI be part of the solution?’

This value-first approach requires pinpointing specific business problems where AI can create a distinct, repeatable financial or competitive edge. The 5% of companies that do see success with AI tend to focus on one critical ‘pain point’ and execute well. AI must be treated as a brand-new nervous system for the enterprise, not a digital assistant.

People, culture and governance

Crucially, this value-first model places a heavy emphasis on building a ‘human-in-the-loop’ culture. Adopting AI brings massive human anxiety, and to combat burnout and distrust, HR departments must lead structured change management programmes. The goal is not to replace workers but to upskill them into ‘supermanagers’ who know how to accurately prompt, challenge, and audit AI outputs rather than blindly trusting them.

This process of co-design, where practitioners’ feedback is routed directly into platform priorities, builds the ownership and confidence that are essential for success.

Establishing structural guarantees and governance is non-negotiable. Rushing into AI without clear rules exposes a firm to severe legal and reputational risks, including cybersecurity incidents that could implicate personal data and copyright infringement claims that have already cost AI developers billions.

Tech leaders face mounting pressure over data privacy due to shifting global data laws, making it imperative that boards set rigid stop conditions, clear escalation paths, and dedicated AI coordination teams to oversee model accuracy, data sovereignty, and compliance. The path forward is not about shunning AI, but about engaging with it strategically and responsibly, ensuring that every investment is a step toward real, measurable value creation rather than a costly leap onto a passing bandwagon.

Peter Bristow is the Change, Transformation & AI Leadership Search Partner at Executive Recruit. This is an exclusive contribution for The Executive Magazine.