Securys is a global privacy consultancy and so we spend a lot of time and energy assessing and discussing the online privacy stances of our clients and their partners and suppliers worldwide. This is time consuming and laborious, but also very helpful, especially when it allows us to look at various companies in a sector and get a sense of how they stack up against each other.  

So what if we could get AI to help? This kind of work is very hard to automate with traditional digital tools as much of it involves complex and subtle semantic analysis of multiple documents. But large language models are, in theory, good at this kind of thing. Could they do some of the leg work for us, and save the Securys team some time which it could use better on more client-facing tasks? And could we set it up in a compliant way, following the best practices in data protection and AI governance that we advise our clients to use? 

To answer this, I set out to build an app. And, of course, I didn’t just want the tool to use AI; I wanted to use AI to help with the programming. I’m a relatively capable hobbyist coder, built a natural-language processing model back before LLMs were a thing, have run multiple dev teams on various projects in the past, and  know my way around a web app. But building a complete application myself? That was something I wouldn’t be capable of doing without about six months of dedicated time and no distractions, and even then it probably wouldn’t be great.  

I did feel I knew enough, though, to put a vibe coding tool to good use. So, after a look around at what was on offer, I signed up with Replit and got to work. And… it’s a long time since I’ve had so much fun with a computer. After a couple of false starts followed by some encouraging experiments I began work on a fully-fledged Privacy Benchmarker tool that we’ve been using in-house over the summer and have sent live for public access on the Securys website today.  

It’s free to sign up and the app will give you six shots at comparing the online privacy stances of companies or organisations by analysing their privacy and cookie notices, looking at their cookie deployment and preference options, and searching the web for records of data breaches and regulatory enforcement actions.  

It bundles the outcomes up into a three-part report that you can cross-compare on a radar chart or export as a PDF. It’s not perfect, but we think it’s quite useful as a starting point when you’re trying to get a sense of how your organisation looks from the outside and how it stands up against other companies in your sector or indeed other organisations whose privacy stance you respect. Have a play and let us know what you think; it’s all very much a work in progress and we’re keen to improve it on the basis of user feedback, so get in touch if you have any useful insights.  

After all, if AI can’t help make data protection efforts less laborious and more effective, how great can it actually be?  

Ready to get started? 

Click here to access our privacy benchmarker tool.