Construction News recently reported that Professional Indemnity insurers are now quizzing firms about AI oversight – with concerns over governance, IP, data security, and negligence risks.

This is the reality of the situation: even if you’ve banned AI, employees are often using it anyway – on their own devices, with unapproved tools.

This ‘shadow AI’ use is rife with risk: customer data, financial details, and commercially sensitive information are routinely being entered into public AI platforms. I’ve seen cases where people uploaded bank statements into free AI tools!

Most employees (around 87%) still have had no AI training, so they don’t understand where that data goes, who has access to it, or the long-term implications.

Remember the recent case of ChatGPT shared links being indexed by Google? That’s the sort of governance blind spot you face without proper guardrails.

Banning AI isn’t a solution – it just drives it underground. The only real answer is to:
~ Upskill leadership teams and staff so they understand the risks and responsibilities.
~Implement an AI policy with clear rules on tool selection, data handling, and approvals.
~Put a robust process in place to vet AI tools properly for enterprise use before adoption.

Whether AI use is official or not, it already presents a governance risk your insurers, and your clients, will care about. The companies that get ahead of this will be the ones with safer operations, lower exposure, and stronger reputations.