Executive Insights, Updated 2023

Shadow AI – hidden impact of deploying Artificial Intelligence



The other day I met with a client of mine, and the conversation naturally fell on the impact of artificial intelligence. The central part of the discussion focused on the state of AI model implementation in organizations today.

Current research indicates that the average number of AI models deployed in organizations today is about five. Still, this figure is expected to experience a 60% CAGR growth over the next four years. As a result, an average large organization will have 35 models deployed in four years, expanding the reach of artificial intelligence. However, this is just the tip of the iceberg, as several other factors substantially impact organizations.

Research conducted by Gartner and McKinsey mainly focuses on officially known AI models. Nevertheless, it’s highly likely that we’ll witness a surge in the deployment of “Shadow AI” models – not unlike the historical spread of Shadow IT (end-user-purchased PCs and software). The rise of AI will trigger a similar pattern as departments like Marketing, Supply Chain, and HR take it upon themselves to deploy AI solutions. This widespread adoption will cause the proliferation of a shadow AI model environment.

Moreover, as AI models are developed and enhanced, their complexity will grow. This complexity refers to the intricacy of tasks or demands managed by the model. It’s likely that a model will also experience an expansion in scope, tackling a more diverse range of tasks typically performed by an individual or product/service. Furthermore, it’s quite probable that an AI model implemented in 2020 would progressively cater to more users each year. As these models mature, their deployment may become economically and technically more feasible for an increasing number of users and business functions.

Effectively managing AI within an organization goes beyond simply dealing with a higher number of deployed models in the coming years. It also entails overseeing model complexity, scope, user count, and the proliferation of Shadow AI. A comprehensive enterprise approach and well-thought-out AI Governance strategy are vital to tackle this burgeoning landscape of AI and Shadow AI. Over time, artificial intelligence will demand increasingly intricate procedures and governance structures for optimal management.

So, what are three things that the organizations need to consider?

Shadow AI

AI Management

As AI models are deployed more extensively in organizations, how will they effectively manage this surge? Furthermore, how will they address the expected rise of Shadow AI?

AI Governance

What does AI governance mean to the organization, who holds accountability for it, and what responsibilities do these individuals carry?

AI Communications

How does an organization foster open communication about AI, instill a sense of accountability among its members, encourage learning AI-related skills for all, and, most importantly, tackle Shadow AI?

In a nutshell

The intricacies of artificial intelligence extend well beyond merely increasing the number of deployed models. It’s crucial for CxOs to stay vigilant and zero in on the potential impacts of Shadow AI on their organizations.

Peter Soendergaard

Peter Soendergaard


Peter is Chairman of the Board at 2021.AI and the Owner of Executive Advisory company, the Sondergaard Group. Before this, Peter worked as the Executive VP and member of Gartner’s operating committee for 15 years. Peter is a well-known and sought out speaker covering many topics within IT, AI & ML.

You might also like…

How to overcome the challenges of productionizing AI

Putting AI models in production is notoriously tricky, and the challenges have many different nuances…


The need for explainable AI

Most blogs, papers, and articles within the field of AI start by explaining what AI is. I will refer to AI as a statistical model which will recognize…