// FEATURE //

lobal AI spend is set to reach

US $ 2.59 trillion in 2026, up 47 % year-over-year. In other words, if AI investment were an economy, it would be the 11th largest – ahead of Canada( US $ 2.51 trillion) and Australia( US $ 2.12 trillion).

It’ s telling of how seriously organisations are taking it. Deloitte’ s latest tech value survey further confirms this, with 74 % of surveyed organisations having invested in AI and Gen AI in the past 12 months – the highest among 20 tracked technologies.

No company wants to be left behind. However, the concern with any rush towards innovation is that companies will be trying to run before they can walk.

In the case of AI, it’ s important to remember that tools will only be as effective as the information that informs it, yet many companies are building AI within environments full of poorly managed, unclassified and outdated files and data.

It’ s an uncomfortable truth, but it needs saying. Without getting your data in order, you could end up creating AI on fragile foundations that can do more harm than good, exposing years of accumulated personal and sensitive information to systems that may not be able to comprehend context or confidentiality.

The roots of IT haven’ t changed

While AI is the shiny new tool of the moment, it is still based on the same fundamentals at the heart of IT: input, process and output.

In both cases, you can’ t mask a poor input with a clever model. If your data isn’ t good data – controlled, risk assessed and properly categorised – then you will be opening yourself up to both failed outputs and information risk before your AI project has even started.

I often say that if someone gave me £ 100,000 to spend on an AI project, I’ d feel comfortable

IN THE CASE OF AI, IT’ S IMPORTANT TO REMEMBER THAT TOOLS WILL ONLY BE AS EFFECTIVE AS THE INFORMATION THAT INFORMS IT.

Intelligent SME. tech

25