AI in Business Report

Artificial Intelligence?On The Ground

Artificial Intelligence (AI) in the 2020s looks more and more like the internet did in the 1990s. Every day, it looks less like a fringe topic for the technically minded and more like one businesses of all types cannot ignore.?

A Google search for ‘AI business’ brings up 1.5bn hits. In March 2021, the UK Government unveiled an ambitious plan to “unleash the transformational power of Artificial Intelligence”. Globally, private investment in AI companies is rocketing. From $8bn in 2015, it reached more than $42bn in 2020. Meanwhile, the EU has set out plans to govern AI , in an attempt to put European values at the heart of the fast-developing technology. But what difference is AI making on the ground today, in the 1.4 million businesses in the UK with at least one employee? These range from tiny two-person firms to Tesco and HSBC.?

How many are actually using AI technology? And in particular, at what stage are most businesses in the AI journey? These are the questions this report sets out to answer. To do so, we need to define AI. There can be a misunderstanding of what AI is. At its broadest, it can be a simple set of rules to automate a process - provided that process involves actions we associate with human intelligence, like perception.

On this definition, a list of instructions to extract the postcodes from a set of addresses could be considered as exhibiting intelligence, although nowadays we may use machine learning to automatically learn the list of instructions. This is the definition we have used here: using computers or other machines for processes that would usually take human intelligence. Within that broad field come more specialised areas like Machine Learning, which uses self-teaching algorithms, in order to find patterns in data and predict likely future trends. It is important to make this distinction because while Machine Learning remains relatively complex there is a huge range of core business processes that could be improved using AI. It is not all about multi-million-pound projects.?

Executive Summary

Peak Indicators engaged the research consultancy Opinium to survey a representative sample of 500 senior decision-makers from UK-based companies about their business’s adoption of Artificial Intelligence in February 2021.

The research discovered that around one UK firm in 30 has started using AI in at least part of their organisation. This muted figure is largely explained by lower rates of adoption in micro-sized firms.

A much bigger proportion of firms have reached the stage of at least testing AI (one in eight), while four in ten have reached at least the stage of investigating its benefits.

There is an AI divide opening up between London and other parts of the UK when it comes to adopting AI.

The UK’s bigger firms are very much more likely to be advanced in AI than smaller ones. Among smaller enterprises, we found an implementation gap. There is an interest in AI and its benefits which is not yet translating into AI projects getting off the ground.

Across all sizes of enterprise, ensuring data quality to allow sophisticated analysis is a key challenge.

Peak Indicators identified three gaps companies are struggling to bridge in the journey towards AI:

1. Going from talking about AI to actually deciding what you can do with it and starting test exercises. The keys here are deeper understanding of what AI is and is not, and then defining the business case for a trial. Key business figures will want to know how long it’s going to take and how much money they’re going to save.
2. Getting data quality right. Only a small proportion of UK firms have reached the stage of data quality that allows advanced AI applications like Machine Learning. The proportion varies with size, from seven per cent of micro firms to 16 per cent of large companies.
3. Going from testing AI to taking the leap of faith to deploy. This can be because initiatives struggle to get full business engagement. The challenge here is cultural more than technical.