I work in a company that heavily utilizes AI, but we're beginning to tighten our controls on which AI vendors we allow. I'd love to hear how other companies manage this process. What strategies do you use to ensure that only approved AI tools are accessible? It feels like a constant struggle to block new tools that emerge and the various URL variations they come with. For instance, how do you permit access to ChatGPT Enterprise while restricting the free version? Currently, we're looking at options like restricting browser access, blocking extensions, and potentially limiting network traffic.
5 Answers
Implementing identity governance is crucial as well. Using tools to track who is utilizing which AI applications is a must. We’ve had great success with Zluri, which alerts us if someone tries to access unapproved tools. It all circles back to having a solid enforcement policy in place.
Good point! It really does help to keep everyone in check while promoting the approved tech we want.
Establishing a strict policy can also work wonders. For instance, letting employees know that any use of unapproved sites could lead to serious disciplinary actions, including termination. It’s about protecting everyone, right?
True, but it really needs to be clear what constitutes 'inappropriate' use, especially with tools like CoPilot.
Right, but it does help in discouraging misuse overall.
Filtering AI access through a proxy is another solid method. You can set it up so all generative AI categories are blocked, and only explicitly approved tools are accessible. That way, you're controlling what gets through easily.
Exactly! Our team does just that, and it has worked well. It minimizes risks while allowing approved usage.
One approach could be using corporate web blockers that have an option to filter out AI-related content. Look for a subscription service that lets you whitelist approved tools while blocking everything else. It simplifies the process, letting a dedicated team manage it for you!
Cisco Umbrella is a solid choice—it has a category to block AI tools and you can easily allow specific ones like Copilot. Just be cautious, as blocking certain domains can sometimes get tricky.
Yeah, that’s a good strategy! It’s not perfect, but it usually does the trick.
Blocking tools like Copilot can indeed be tough! Microsoft’s configurations don’t always behave as advertised, and they keep finding ways to push it at users. It’s been a headache for many of us!
For real! They seem completely blind to what users actually want or need when making these updates.
Exactly, it’s about balance—monitoring without being overly restrictive.