You’ve chosen your AI tools. Now comes the bit that probably didn’t feature in the demo: policies, data protection, security and working out what people should actually be allowed to do with them.

In this episode, Adam and George talk through the AI governance Garwood has been putting in place, from what we expect from our technology providers to the policies and controls we need internally.

The challenge is getting the balance right. You need to protect client and company information without putting so much friction around AI that nobody actually uses it.

They get into:

• Why business AI accounts matter when handling client information
• What to check in vendor terms, DPAs and security assurance
• Where your data is stored and processed
• The risks that come with connecting AI to other systems
• DPIAs, legitimate interests and acceptable use policies
• Being transparent with clients about AI
• Why human review still matters

There’s also a cautionary tale involving someone using their company ChatGPT account to redesign their garden and apparently racking up thousands in additional usage fees.

The garden may have been cheaper.

Next time

Adam and George get more hands-on with the AI skills we’ve built at Garwood, including how we’re getting AI-generated documents to look and sound more like us.