Stop the madness! AI is a CEO mandate

Isn’t it interesting that the adoption of artificial intelligence (AI) is soaring, but the results are disastrous. Enterprise investment in GenAI tools is in the billions, yet the failure rate for AI projects is shockingly high, estimated to be over 80%. For custom enterprise-grade GenAI tools, only 5% reach production. This chaos of fragmented projects and failed pilots costs organizations millions, with estimates suggesting between $800,000 to $200 million is wasted per successful project.

The root of this widespread problem isn’t a technology issue; it's a leadership and organizational failure. We consistently work with teams struggling to secure management buy-in for AI projects, leading many organizations to prematurely adopt AI through IT, with disappointing results.

So foundational and fundamental is the effect of AI that Effective AI adoption demands top-down leadership and strategic mobilization at the executive level. It’s my opinion that to stop this financial and organizational madness, AI must transition from being isolated technical projects often driven by vendor renewals or license deals to a core strategic mandate driven by the CEO.

I. The Root Causes of Failure: Why Outsourcing AI to IT Fails

The high failure rate stems from treating AI as a standard technological deployment rather than a complex, cross-functional organizational transformation. This flawed perspective creates a series of costly problems:

• Strategic Misalignment and Fragmentation: Leaders often operate with an anecdotal, need-based approach to AI when they should be taking a holistic, "all-in" approach. Projects are launched without strategic planning and most often driven by a vendor agenda, leading to uncoordinated initiatives, duplicated work, and internal silos. Without a unified vision, fragmentation is inevitable, which results in organizations incurring much higher technology expenses and achieving dramatically less impact compared to those with a unified AI strategy.
• The IT Trap: Analytically immature companies frequently treat complex AI projects like IT initiatives. They focus heavily on the technical aspects and give responsibility to IT teams. This fundamental mistake ignores the crucial human elements and change management required.
• Lack of Executive Authority: Unbelievably the lack of support or buy-in from leadership is one of the most persistent issues. Without strong executive sponsorship, which is cited as the absolute first requirement, the initiative lacks the necessary budget, authority, and organizational credibility to enforce the levels of organizational transformation or the adoption of standards to drive the change.
• The Learning Gap: The primary factor keeping organizations trapped is that most AI tools deployed do not retain feedback, adapt to context, or improve over time. This learning deficit is critical for mission-critical work. Users often prefer simple consumer tools (you will have heard of the war stories of leaked data) because they are more familiar and responsive, but they abandon both internal and external static enterprise AI solutions because they require too much manual context input each time and fail to learn from previous interactions.
• The Vendor agenda: Vendor’s striving for growth land not only their tools, but tool training, teaching people about the tools, versus teaching people how to work and think with AI. The approach should rather be: how do we use AI tools to help people to think better and accomplish super-human capabilities?

II. The Alternative: The CEO AI Business Priorities Framework Mandate

Anyone who has ever worked in a matrix management structure knows only too well that to get something done across the business you create a cross-functional team of SVP’s into a Center of Excellence. Right. Wrong! That is not what I am suggesting at all.

The C-Suite must be the AI Center of Excellence headed up by the CEO establishing an organizational structure that engages experts to advise, guide, and oversee AI transformation.

This model must be established to explicitly bridge the gap between executive decision-making and AI implementation.

Why the CEO Must Own the Mandate:

Our business was founded on the idea that we had seen the future and wanted to be in it. A thought provoking report we provide to our clients is this, “We’ve seen the future. You are not in it”.

The COE is accountable for the success or failure of a business. Who can forget the alleged famous words of then Nokia CEO Stephen Elop who said, “We didn’t do anything wrong, but somehow we lost” at the collapse of Nokia.

Given the tectonic shift taking place, the first order of business as I see it is the CEO taking on the role of Chief AI Officer, they are the only person with the authority to make the decisions that will be required. It is madness to outsource such a critical function to anyone let alone the IT Department to pilot AI with some basic use cases?

It’s clear this is not a technical decision. We already know the technology works. It’s now about how you play this game. How quickly you are learning, how quickly you adopt the insights and how quickly you transform to executing these insights with accountability structures enforced from the highest level?

Here is the outcome we are driving for.

1. Executive Education: Establishing a common foundation of understanding is critical for the C-Suite. Executives need to know how to use the most advanced AI models to think and to accomplish leverage for strategy, key decisions, and addressing the most complex of business problems. In this regard it is obvious why current vendor training is totally inadequate for this requirement.
2. CEO AI Business Priorities Framework: What is the Strategy and Futures Perspective across the C-Suite at a functional level? Without a clear and informed understanding of the impacts, and more importantly the opportunities created by AI, how is each executive going to plan and respond. The AI Business Priorities Framework sets forth these priorities and plans.
3. Organizational Education: Organizational transformation will be an outcome of adopting AI projects and initiatives from the AI Business Priorities Framework. Making sure that the organization understands AI in the same way as the executive leader is critical for 2 reasons.
   1. To ally the fears any AI project elicits.
   2. To ensure that staff at all levels are leveraging AI to maximum advantage for the business itself. This goes well beyond being able to use AI to respond to emails.
4. Securing Authority: Seeking approval from C-suite sponsorship to provide the strategic direction and resources needed for decisions are sped up, standards can be adopted to drive organizational change quickly. This commitment must be visible and long-term to overcome resistance and funding challenges.
5. Enforcing Governance: Accountability is a core feature of the mandate. The increasing complexity of AI governance demands centralized oversight. In fact, 28% of organizations assign AI governance to their CEO, a practice linked to higher bottom-line impact.

III. How the CAIO CoE Stops the Madness (Why the Alternative Works)

This approach replaces chaotic, fragmented adoption with a structured AI Business Priorities Framework built on clear strategic alignment, stringent governance, and standardized execution.

1. Strategic Alignment (Ensuring Investment is Focused)

Informed by the strategic insights in the Business Priorities Framework the CoE’s foundational responsibility is to execute on the AI strategy that is aligned to the business goals. This is how it ensures value delivery:

• Prioritizing High-Value Use Cases: The CoE establishes a systematic approach for identifying and prioritizing business use cases. It ranks ideas based on complexity and value criteria, seeking "quick wins" (High potential/Low complexity) to prove AI’s value early and build credibility.
• Defining Measurable Outcomes: The CoE requires setting clear decision-making authority and defining KPIs (Key Performance Indicators) such as ROI and adoption rates to ensure that AI projects are directly tied to revenue growth, cost reduction, or efficiency improvements.
• It aligns Prioritized areas for innovation and disruption: Testing your strategy against emerging Metatrend with Causal and Impact Analysis is now a possibility. Using AI to tease out insights before the markets ensure you have a spot in the future.

2. Standardization and Efficiency (Eliminating Waste)

The CoE acts as a central repository for expertise and resources, accelerating adoption and deployment by eliminating redundancy:

• Standardized Practices: It develops a set of standardized practices and processes for AI development, deployment, and lifecycle management. This standardization is essential for making scaling AI efforts easier.
• Infrastructure and Data: The CoE designs AI-ready infrastructure and provides access to shared resources like cloud platforms and machine learning frameworks. Critically, it establishes a common data model to streamline data integration and ensure AI systems can access accurate, current data from relevant platforms efficiently.

3. Governance and Ethics (Mitigating Enterprise Risk)

The CoE is essential for managing the new layer of risks introduced by AI:

• Mitigating Risk: The CoE provides oversight by establishing governance frameworks, policies, and processes. This ensures projects adhere to ethical and regulatory standards, covering risks such as fairness, data integrity, privacy, and accountability.
• Protecting Reputation: Without clear governance, organizations risk compliance failures, reputational damage, and diminished public trust. The CoE helps mitigate these factors by monitoring models for bias and ensuring compliance with regulations like GDPR or HIPAA.

4. The Optimal Structure (Balancing Control and Speed)

To deliver value effectively, the CoE should adopt a flexible structure that prevents it from becoming a central bottleneck or an echo chamber.

• Tightly aligned and informed with the CEO and C-Suite: Speed of execution is the differentiator. How do you turn today’s conversations into tomorrow's prototypes and next month's products? You have to totally re-imagine the enterprise innovation pipeline if one exists. Today’s approach will not solve tomorrow's requirements.
• Federated (Hybrid) Model: The federated model is often the most effective structure. It integrates centralized strategy and governance (the Hub) with decentralized execution (the Spokes).
• Why it Works: The central CoE adds the execution and implementation plans to the strategy, standards, and ethical guardrails, while most AI talent remains embedded within business units to customize and deploy AI for business-specific use cases. This structure achieves central strategic alignment combined with the business-unit speed and ownership needed for innovation.

Conclusion: The Window is Closing

The high failure rate of AI projects is not a technology problem; it is an organizational failure rooted in a lack of strategic alignment and executive accountability. The approach outlined is one that we have embraced and are working with a number of local and international businesses to explore its implications and benefits. In the Age of AI you have to create a learning organization that outperforms the competition through AI leverage.

If being able to read and analyze an Excel spreadsheet is the skill of today’s C-Suite, it’s our view that in the not too distant future that will become the latest program you created whilst analyzing the last months board meeting notes.