Technical debt is hindering AI development

There’s a lot of talk about how AI can solve challenges and create value in Norwegian businesses – from automating routine tasks to advanced functions like using generative AI to plan maintenance of power grids. However, a key requirement for this is access to large amounts of reliable, company-specific data. It is here that, companies’ technical debt creates a significant obstacle.

According to Gartner, technical debt is expected to exceed $1 trillion globally by 2025, highlighting how widespread the problem is, not just in Norway but worldwide.

Technical debt can be compared to a house from the 60s or 70s, where there have only been superficial upgrades. Poor ventilation, outdated plumbing, and electrical systems make it difficult to implement new technologies like smart home solutions. For example, if someone wanted to install solar panels, they would first need to upgrade the house’s entire electrical system.

Unfortunately, this scenario is also typical of Norwegian businesses. Many companies created before 2000 still have IT solutions from the 80s and 90s, which have been maintained but not modernised. New IT solutions have been layered on top of the old systems, often with quick fixes to make them work together. Additionally, many new systems were developed rapidly to meet market needs but they have poor quality code, a lack of testing, and weak integration with the rest of the IT architecture.

This makes up companies’ technical debt: a mix of old legacy systems and new solutions that were created quickly and/or don’t work well together.

The consequences of this accumulated technical debt are that companies are unable to fully harness the potential of new tools like AI and automation. Simply put, AI relies on large amounts of data, and the quality of that data must be high. For example, if AI is to be used for decision support by a management team, it’s crucial that they trust the analyses and recommendations provided by the AI model.

High technical debt also increases the risk that data storage and management processes are not compliant with current regulations such as GDPR, NIS2, etc. This means there is a greater likelihood of serious security incidents or fines from regulatory authorities.

By addressing technical debt and modernising older IT systems, companies will secure better data control and higher data quality, both of which are crucial for successful use of artificial intelligence. In other words, Norwegian companies need to start with a proper upgrade of their digital “house”!

Read the article in Norwegian.