Alright folks, listen up! OpenAI just launched their ChatGPT Enterprise, and they’re really trying to address all those privacy concerns and fears that business customers might have. So, what exactly does this enterprise-level access get you?
Now, before we dive in, let’s get a few things straight. OpenAI made it clear that “all conversations are encrypted in transit and at rest.” And this applies to everyone using ChatGPT, whether you’re on the free tier or the enterprise tier, whether you’re using the API or the bot interface. It’s all encrypted using HTTPS or specifically, TLS 1.2.
In terms of conversations being encrypted at rest, OpenAI uses the good ol’ AES-256 algorithm to secure customer data in storage. But here’s the thing, they told The Register that even non-enterprise chats are encrypted too.
In their FAQ, OpenAI also mentioned that for enterprise customers, certain staff members have the ability to unlock and access those encrypted business conversations when necessary. So, there’s that added layer of accessibility.
Now, OpenAI wants to emphasize that your precious business data and conversations aren’t used to train ChatGPT’s model. They explicitly stated that “We do not train on your business data or conversations, and our models don’t learn from your usage.”
However, here’s something worth noting. According to OpenAI, data submitted through the API isn’t used to train or improve ChatGPT for any tier, enterprise or not. But for those chatting away with the bot outside of the API, the provided data might be used for training purposes. Of course, you can choose to opt out of that if you want.
Right now, it seems like the enterprise tier offers the same level of encryption as the non-enterprise tier. Data privacy via the API is also on the same level for both. However, there’s one key difference. By default, the chat bot interface doesn’t train the model from conversations for the enterprise tier, while regular users have to opt out of that.
OpenAI really needs to address these privacy and security concerns because big players like Apple, Amazon, Samsung, JPMorgan Chase, and Bank of America have banned their employees from using ChatGPT. They’re worried that sensitive or proprietary data might accidentally find its way into the model and end up leaked to third parties.
But OpenAI wants to put all those worries to rest. They want you to know that “You own and control your business data in ChatGPT Enterprise.” It’s your data, your control. Simple as that. They also want to remind us that ChatGPT is not a flawless bot. It might sound confident, but it can still produce inaccurate results.
So, what’s the big deal with the enterprise tier then? Well, for starters, you get some fancy administrative controls. They’ll give you a console to manage all your ChatGPT users in one place. Single-sign-on authentication? Check. User domain verification? Check. And they even throw in an analytics dashboard to monitor bot usage. So, you’ll have all this control and oversight.
But wait, there’s more! Enterprise customers get unlimited access to a GPT-4-powered ChatGPT. And here’s the kicker: it runs up to twice as fast on OpenAI’s servers compared to the non-enterprise GPT-4 usage. That’s some serious speed right there. And on top of that, you also get access to their Code Interpreter data-analysis tools. So, plenty of firepower for those enterprise folks.
Here’s another perk for enterprise users. You can use these 32,000-token context windows. Tokens basically measure and limit the size of input data. One token represents about four