OpenAI Launches ChatGPT Enterprise With Unlimited GPT-4

OpenAI unveiled ChatGPT Enterprise today, a version of the chatbot built specifically for businesses that want to use it without the usage caps or privacy concerns of the consumer version. The announcement comes four months after the company opened a waitlist for "ChatGPT Business," the provisional name under which the product had been known since April.

The core promise is easy to grasp: businesses will be able to use GPT-4 without the usage limits that ChatGPT Plus subscribers face, with faster response times, and with a guarantee that no conversation or data entered into the platform will be used to train OpenAI's models.

What the corporate version includes

ChatGPT Enterprise adds a layer of security that didn't previously exist in ChatGPT: encryption of data in transit and at rest, compliance with the SOC 2 standard (a widely required certification that companies demand from their software vendors to verify they handle information securely), and admin tools such as single sign-on (SSO) and domain verification for centralized management of employee accounts.

The context window — the amount of text the model can "remember" and process within a single conversation — is four times larger than in the standard version, reaching 32,000 tokens. That allows users to work with lengthy documents, extensive reports or long conversations without the model losing track of the thread.

The product also includes Advanced Data Analysis (formerly known as Code Interpreter), the tool that lets users upload files and ask ChatGPT to analyze data, generate charts or write code to process them — particularly useful for finance, analytics or research departments.

Privacy as a selling point

Since its public launch in November 2022, ChatGPT has run into resistance from many companies reluctant to share sensitive information with a service whose free-tier terms allowed conversations to be used to improve the models. Several companies, including Samsung, banned employees from using ChatGPT after internal leaks of code and confidential documents came to light.

ChatGPT Enterprise addresses that problem head-on. OpenAI guarantees that data from corporate clients won't be used to train its models and that the company itself won't have default access to those conversations unless required for security or regulatory compliance reasons.

A market OpenAI needs to win over

The launch comes as OpenAI competes with Microsoft, which builds GPT-4 technology into Copilot for Microsoft 365, and with Google, which is pushing Bard among its Google Workspace customers. What sets ChatGPT Enterprise apart is that it offers the model more directly, without requiring companies to already be using Microsoft's ecosystem or another productivity provider.

OpenAI hasn't disclosed pricing for the service, which is negotiated case by case depending on each client's size and needs — a common practice in enterprise software, but one that contrasts with the flat $20-a-month rate ChatGPT Plus charges individual users.

What it means for the industry

The move confirms that OpenAI, beyond being a research lab, has become a software company with a growing B2B business model. The company needs recurring revenue from corporate clients to sustain the cost of training and running models like GPT-4, which require extremely expensive computing infrastructure.

For businesses, ChatGPT Enterprise removes much of the technical excuse for not adopting the tool: there's no longer a need to choose between productivity and data control. What remains to be seen is whether the still-undisclosed pricing will be accessible to mid-sized businesses, or whether the product will end up reserved for large corporations with software budgets that match their AI ambitions.