Skip to content

Three mistakes to avoid when implementing AI 

Because of high expectations, decisions to deploy AI are often made based on incorrect assumptions. Lassi Kurkijärvi and Mia Nikula from Sofigate give their advice on how to avoid the most common mistakes. 

Right now, one topic above all others is dominating the conversation in companies: Artificial Intelligence. The promises of AI increasing efficiency drastically and rapidly are so high that management bodies within companies feel immense pressure to act quickly. With the tension between enthusiasm and expectations, decisions are often made based on vague or even incorrect assumptions. 

But these things should not be rushed. Projects implemented in a hurry rarely produce the desired effect on efficiency. Transformation based on wrong assumptions can also have unexpected consequences that are undesirable or even negative. 

Therefore, it is better to correct at least the following misconceptions before embarking on a journey to implement AI. 

1. The introduction of AI is an IT project 

AI uses technology, so surely its implementation is an IT project? The project should be managed by the IT team, using best practices for IT projects and starting with quarterly planning, right? 

Wrong. True business transformation should always be the responsibility of the company management.  

Only company management can ensure that transformation is carried out on the terms of the business and not the technology used. It is also the responsibility of the company management to ensure that change does not occur in silos. Instead, the whole organisation should be involved, each individual and unit in their own way. 

Traditional IT process models for design and management are helplessly slow for AI projects. At the same time, AI radically accelerates value creation and contributes to shorter planning and implementation cycles. 

2. Our organisation is so unique that it requires tailor-made solutions 

Many organisations believe that their unique business and operational challenges can be solved only with the latest technology available. Based on this belief, they want solutions that are tailored to their business needs but often expensive and slow to implement. However, these projects often deliver only pseudo-solutions that improve isolated functions without addressing the core of the business and value creation. 

Most organisations are already using modern business platforms that include most of the technology they need, including AI. Based on experiential learning, simultaneous top-down and bottom-up change enables the agile introduction of new capabilities. This approach yields results that serve both the day-to-day needs of employees and the core strategic objectives of the business. The Business Technology Standard also provides ready-made business models, which means you don’t have to completely reinvent the wheel, even for AI projects. 

This is another reason why it is important to collaborate and work across silos. Company management needs to be familiar with the business technology in order to understand the opportunities it offers. IT management, in turn, needs to understand the business objectives that the new solutions are designed to serve. 

3. Involving the whole organisation in the change is too expensive 

Major transformation projects require the organisation to make significant investments in terms of both time and human resources, so the actual management and implementation of the project are often outsourced to consultants. As a result, however, the organisation does not actually increase its change capacity at all. This means that in order to build upon the foundation of the newly implemented change, the organisation needs to bring in yet another army of consultants. 

Even if outsourcing produces savings in the short term, the costs quickly add up. Increasing the organisation’s own capacity for change is therefore a far more profitable investment. 

Building up the in-house change capability is particularly important for AI projects since they are implemented quickly and cover every area of the operation. In order to overcome any possible resistance to change and to make sure that the new ways of working are embedded in the corporate culture, every member of the organisation must be involved in the change process. 

Therefore, having everyone from the organisation participate is not too expensive. The far more costly mistake would be not to include everyone.  


AI will not replace the worker – it makes them irreplaceable 


Lassi Kurkijärvi has been building digital businesses for a quarter of a century. Today, he helps Finnish companies in digital transformation in his role as Executive CTO at Sofigate. His passion is humanity in all its forms: change is ignited by inspiration and engagement – and extinguished by the lack of them. 

Mia Nikula has had many roles in major business transformations both as a manager and as a senior expert. She currently acts as Business Executive for Platform Transformations at Sofigate. She has a passion for growing talent in business and especially leading people through a major change. Her motto is: “Inspire, lead by example, enable, bring out the benefits and engage in continuous dialogue.”