The rise of shadow AI in small businesses

Artificial intelligence is rapidly becoming a normal part of working life, but for many small businesses, its adoption isn’t necessarily happening through a formal strategy or company-wide rollout. 

According to the Office for National Statistics (ONS), 55% of employees report using AI for work or education. However, just 35% of businesses report using at least one AI technology. In other words, AI use among employees appears to be outpacing formal adoption at a business level.

This gap between employee use and formal business adoption is increasingly being described as “shadow AI” – a situation where employees use artificial intelligence as part of their work without a clear organisational approach for managing it.

The fact that employees are using AI independently isn’t necessarily a cause for concern. However, without visibility over how these tools are being used, businesses will struggle to ensure that AI is being used safely, consistently and entirely effectively.

Why recognising shadow AI matters

AI is becoming ever more popular because it can help employees complete tasks more efficiently. The right tools can support activities such as drafting content, summarising information, generating ideas and automating repetitive work. This can be especially valuable for those working in smaller organisations, where there’s often a need to juggle multiple tasks and responsibilities at once.

However, AI use can create challenges if businesses don’t have visibility over how it’s being utilised across the organisation. Without knowing which tools employees are using and what role it’s playing in day-to-day work, it can be incredibly difficult to identify where (and when) safeguards are needed around data protection, accuracy and human oversight.

It’s becoming increasingly clear that businesses need to develop a clearer understanding of what good AI use looks like. This means identifying where the technology can genuinely improve the way people work, while also ensuring employees have the knowledge and support to use it effectively.

Bringing AI use into the open

For most small businesses, the first step in managing shadow AI is going to be understanding how it’s already being used. Rather than assuming employees are using AI responsibly, or choosing to ban its use altogether, business owners are likely to benefit from having open conversations about where and why AI tools are being embraced.

This could involve asking employees which tools they use, what tasks they use them for, where they believe AI is helping them work more effectively, and where there are still gaps. By building a clearer picture of current behaviours, businesses can identify where AI is adding value, and where additional guidance may be needed.

Creating an internal approach to AI doesn’t necessarily mean introducing complex policies or restricting access. For many small businesses, it could be as simple as putting a few basic principles in place, such as:

• Which AI tools employees are encouraged or approved to use (and why)

• What information should never be shared with AI systems

• When AI-generated work should be reviewed by a person

• Providing basic training on AI's strengths and limitations

Simple, clear guardrails can help employees use AI with greater confidence, while also giving businesses greater visibility over how it’s being used.

Alongside internal policies and training, it may also be worth reviewing your business insurance arrangements to ensure they’re still appropriate. While insurance won't replace good AI governance, it can form part of a wider risk management strategy.

The bottom line

Shadow AI isn’t necessarily a sign that employees are using technology irresponsibly, but it does highlight how quickly AI is becoming part of everyday work, often before businesses have had time to formally assess the best approach.

For SMEs, a clearer understanding of how AI is already being used can help identify where the technology can add value, and where additional consideration may be required. By creating clear guidance, supporting employees with training, and establishing where human oversight is needed, it’s possible for AI to be used effectively, and to create an environment where innovation and responsibility can go hand in hand.