Why Is the Use of AI Still Limited in Finnish Companies?
- Markus Mertanen
- Sep 1
- 2 min read
Updated: Sep 2
When preparing Aines, we wanted to accurately understand the current state of AI utilization in Finland. Over the summer, we interviewed decision-makers from more than 20 Finnish companies.
The findings from our market research are clear and align with international studies: Generative AI is on the investment agenda in almost every company, but established solutions that have been widely adopted and deliver significant business value remain very limited.
Why are the promises of AI not yet fully visible in the daily operations of companies? Based on the interviews and our own experiences, here are the key reasons we've identified:
1. Experiments remain isolated
Companies are actively launching various AI pilots. The goal is often to build some sort of an "internal ChatGPT" for a specific business purpose.
However, most experiments are discontinued or reduced to small-scale testing without business impact, and solutions are rarely taken into production. Usage of company-specific chatbots tends to be low, as employees prefer familiar general-purpose AI tools.
Key challenges in conducting AI pilots include the lack of in-house expertise, limited time available, and unclear responsibilities related to AI utilization, deployment, and maintenance..
2. Lack of understanding of generative AI
Many companies still do not fully grasp what generative AI can truly enable.
Thinking is often based on the business processes and limitations of existing IT systems, and AI is not yet seen as an integral part of core business workflows.
AI is frequently viewed merely as a chatbot, requiring active interaction and new learning from employees. This leads to a situation where the benefits of AI depend heavily on individual employees' skills and attitudes.
Many companies still do not fully grasp what generative AI can truly enable.
The challenges inherent to generative artificial intelligence, such as hallucination and randomness, also create distrust and hinder its use in operational business activities.
3. Technological challenges
Generative artificial intelligence is developing rapidly, and interviewees found it difficult to decide which technology or language model to build on for the long term.
Implementing reliable solutions that leverage a company’s own data requires specialized expertise, such as prompt engineering, understanding of large language models, and software development skills suited for AI solutions. According to the interviews, companies almost always need to acquire this competitive expertise externally.
The business data needed for AI solutions is not easily accessible and is often scattered across different information systems, data warehouses, emails and shared files. Additionally, data security and the security of AI solutions were perceived as unclear.
How to get from experiments to concrete results?
Based on our market research, companies have a strong desire and need to integrate AI into their own core business workflows, but due to the challenges we have summarized, they need the help of the right partner and production-ready solutions that can be adapted to them.
