Tim DaviesCEO Projekt Rising - The expert AI partner for SMEs

Are AI Chatbots Leaking Your Company's Secrets?

20 Aug 2025

Watch on YouTube

Tim Davies warns that employees commonly use large-language-model chatbots with company data outside formal AI policies, giving an example where a company banned ChatGPT but employees used it on phones and emailed outputs back to work — creating data-leakage risks. He argues businesses need appropriate structures and frameworks to manage how data is used to prevent new avenues of problems.

Now, how many individuals within companies are using large language model based chatbots with company data and doing so outside of their company's AI policy? Yeah, I know of a company where they had a policy, they blanket said thou shalt not use chatgpt. Yeah. So what did their people do? Well, they use it on their phones and they email the output to their work email address and then they're using the output. So they're still using the AI process, they're still leaking data that they shouldn't be. And so that sort of recognition of how data is used within a business and putting in the right structures and framework is absolutely critical. Otherwise, actually, what they're doing is opening up a whole new avenue of problems.