# What is ChatGPT Enterprise, and how is it different from regular ChatGPT?

Date: 2023-08-29 · Author: Robert Scales · https://guidetoartificialintelligence.com/en/learn/what-is-chatgpt-enterprise/

**Answer:** ChatGPT Enterprise is a business version of ChatGPT that OpenAI launched on August 28, 2023, with unlimited and faster GPT-4, a longer memory for documents, data analysis tools and an admin console. Its main difference is a promise about data: OpenAI does not train its models on a company's conversations or business data, and everything is encrypted.

Yesterday OpenAI launched **ChatGPT Enterprise**, a version of its chatbot designed for whole organizations rather than individual users. It looks and works like the ChatGPT millions of people already use: you type a request and get a written answer. What changes is what happens around the conversation, from who controls the data to how much of the most capable model employees can use.

Bloomberg calls it OpenAI's most significant attempt so far to win a broad range of business customers and to earn more from its best-known product. Running large AI models takes enormous computing power, so turning ChatGPT's popularity into steady revenue matters to the company.

## Why businesses asked for it

ChatGPT is only nine months old, yet OpenAI says teams at more than 80% of Fortune 500 companies (the 500 largest US companies by revenue) have adopted it. Many of them did so informally, with employees signing up on their own.

That worried IT and legal departments. On the consumer service, conversations can be used to improve OpenAI's models unless a user opts out, according to Bloomberg. Some companies kept away from chatbots entirely because they feared confidential information, such as customer records, contracts or code, could leak or end up in training data. Business leaders told OpenAI they wanted a simple, safe way to deploy the tool across their staff.

OpenAI says it tested the new product with more than 20 companies, and it names early users including the design platform Canva, the payments company Block, the beauty group Estée Lauder, the investment firm Carlyle, the consultancy PwC and the automation service Zapier.

## What ChatGPT Enterprise includes

The announcement groups the features into three areas.

| Area | What you get | What it means in plain words |
|---|---|---|
| **Privacy and security** | No training on business data or conversations; AES-256 encryption for stored data and TLS 1.2 or later for data in transit; SOC 2 compliance | Your company's chats are not used to teach the model, data is scrambled while stored and while traveling, and an outside auditor has checked the security controls |
| **Administration** | Admin console with bulk member management, single sign-on (SSO), domain verification and a usage dashboard | IT can add or remove staff in one place, employees log in with their work account, and managers can see how the tool is used |
| **Capability** | Unlimited GPT-4 with no usage caps, up to twice as fast; 32,000-token context; unlimited advanced data analysis | The strongest model without hourly limits, room for much longer documents, and a built-in tool that runs code to analyze files |

A few terms deserve a closer look. The **context window** is how much text the model can keep in view at once. At 32,000 **tokens** (pieces of words), it is four times the standard window, so a long report or contract can fit in one conversation. **Advanced data analysis** is the new name for the feature previously called Code Interpreter: you upload a spreadsheet, and ChatGPT writes and runs small programs to summarize, chart or clean the data.

There are two more additions. **Shared chat templates** let colleagues reuse proven prompts for common tasks, and the price includes free credits for OpenAI's API, the paid programming interface, for companies that want to build their own tools.

## How much does it cost?

OpenAI has not published a price. Brad Lightcap, the company's chief operating officer, told Bloomberg that the cost depends on each business's needs and that OpenAI will work out a plan with each customer. For now, interested companies contact OpenAI's sales team, and OpenAI says it is onboarding as many as it can over the next few weeks.

OpenAI also lists features still to come: connecting ChatGPT securely to the applications a company already uses, a self-serve ChatGPT Business plan for smaller teams, more powerful data-analysis and browsing tools for work, and tools designed for specific roles such as marketing or customer support.

## Why it matters

For many organizations, the question has been less "Is ChatGPT useful?" than "Can we allow it?" A clear written promise not to train on business data, plus encryption, audits and central control, answers the most common objections from security and legal teams. That could move ChatGPT from quiet personal use to an official workplace tool.

It does not remove every risk. Language models can still state wrong things confidently. Lightcap points out that GPT-4 makes up information less often than OpenAI's previous model, but less often is not never, and checking important answers remains the user's job.

> **Try it:** List three tasks in your job where you would like help from a chatbot. Next to each, write what information you would need to paste in. Mark any item that is confidential. That simple list shows whether a consumer account is fine for you or whether you need a business plan with data protections.

## What should a company check before rolling out ChatGPT to its staff?

Start with data. Read the provider's terms to confirm whether conversations are used for training, how long they are kept, and where they are stored. Decide which kinds of information employees may paste in, and write that down in a short, plain policy.

Next, look at access and oversight. Single sign-on lets you remove access the day someone leaves; a usage dashboard shows which teams find the tool valuable. Agree on who in the company owns the tool, answers questions and reviews new uses.

Finally, invest in training. Show staff how to give clear instructions, how to check answers against reliable sources, and when not to use AI at all. A secure tool used carelessly can still produce a confident mistake, and the person who sends it is still responsible for it.

## Sources

- Introducing ChatGPT Enterprise: https://openai.com/index/introducing-chatgpt-enterprise/
- OpenAI Unveils ChatGPT for Businesses, Stepping Up Revenue Push: https://www.datacenterknowledge.com/artificial-intelligence/openai-unveils-chatgpt-businesses-stepping-revenue-push