Anthropic is an American AI safety and research company, based in San Francisco, that makes Claude, the AI assistant and family of models. It was founded in 2021 by former OpenAI staff. The detail people miss: it's a public benefit corporation, so it has a stated purpose beyond serving its owners.
If you use Claude, you're using a product made by Anthropic. Plenty of people know the assistant's name without knowing much about the company behind it, and that matters more than it sounds. Who builds an AI tool shapes how it behaves, what it refuses to do and how it treats your data.
Here's a plain-English overview of Anthropic: who started it, what it makes, how it says it approaches safety, and what that means if you're a small business owner choosing an AI assistant.
What Anthropic is
Anthropic is an American AI company headquartered in San Francisco. Wikipedia describes it as a public benefit corporation, and Anthropic calls itself an AI safety and research company. Its best-known product is Claude, a family of AI models and the assistant apps built on them.
A public benefit corporation is a type of company that has a stated purpose beyond making a return for its owners. Anthropic says its purpose is the responsible development of advanced AI for the long-term benefit of humanity.
Who founded Anthropic and when
According to Wikipedia, Anthropic was founded in 2021 by former members of OpenAI. Its leaders are siblings Dario Amodei, the chief executive, and Daniela Amodei, the president, alongside other co-founders who'd worked on AI research together.
The company was set up with a strong focus on AI safety, and that focus still runs through how it describes itself. Amazon and Google are both among its investors, which helps explain how it can train very large models.
What Anthropic makes: Claude and Claude Code
Anthropic's flagship product is Claude, which Wikipedia describes as a series of large language models. Most people meet Claude through the Claude apps on the web, desktop and phone, but businesses can also build Claude into their own products through Anthropic's developer platform.
| Anthropic product | What it's for |
|---|---|
| Claude apps | Chatting, writing, research, analysis and working with files |
| Claude models | The AI models behind the apps, also available to developers |
| Claude Code | A command-line AI agent designed for coding |
| Model Context Protocol | An open standard for connecting AI assistants to tools and data |
If you're deciding between assistants, my comparison of Claude and ChatGPT for business walks through the differences. And if coding is on your list, my guide to using Claude Code when you have never written code is a friendly first step.
How Anthropic approaches AI safety
This is the part of Anthropic's identity it talks about most. On its company page, Anthropic says it aims to build frontier AI systems that are reliable, interpretable and steerable, and that it treats AI safety as a systematic science.
One technique it's known for is Constitutional AI. Wikipedia explains that this trains a model to follow a written set of principles, which Anthropic calls a constitution. The idea is that the model's behaviour is guided by clear values rather than only by examples.
- Reliable: the AI should behave consistently and predictably.
- Interpretable: researchers should be able to understand why it does what it does.
- Steerable: people should be able to guide its behaviour.
How Anthropic is run
Anthropic has an unusual governance set-up. Its company page says its board of directors is elected by its stockholders and by a separate body called the Long-Term Benefit Trust. The trust is designed to keep the company's long-term mission in view alongside ordinary business decisions.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
You don't need to understand corporate structures to use Claude, but it's useful context. When a company says safety is central, its structure is one place to see whether that's built in.
Anthropic's role in the wider AI world
Anthropic is one of a handful of companies building the most advanced AI models, alongside names like OpenAI and Google. It's also contributed shared standards. In late 2024 it introduced the Model Context Protocol, an open standard that lets AI assistants connect to the tools and data businesses already use, and other AI companies have since adopted it.
What Anthropic means for small businesses
For most business owners, Anthropic matters because of Claude. You don't need to follow the company's research papers to get value from its tools, but knowing its priorities helps you understand why Claude behaves the way it does. Here's how I'd think about it:
- Claude is strong at long documents, careful writing and following detailed instructions.
- Anthropic's safety focus means Claude may decline some requests other tools accept.
- Business plans and admin controls exist for teams, so check Anthropic's site for what each includes.
- Read the privacy and data settings in your Claude account before sharing client information.
On that last point, my guide to how Claude memory works and how to turn it off explains one of the settings worth checking. And if you'd like help deciding where Claude fits in your business, that's the kind of work I do as an AI consultant.
Common questions people mix up about Anthropic
Because the AI world moves fast and the names are similar, I see the same few mix-ups again and again. Here are the ones worth clearing up before you choose a tool or explain it to your team:
- The company isn't part of OpenAI, though its founders previously worked there.
- Claude is the product; Anthropic is the company.
- Anthropic doesn't own Amazon or Google, or the other way round; both are investors.
- The Model Context Protocol came from Anthropic but is an open standard anyone can use.
Frequently asked questions
What does Anthropic do?
Anthropic is an AI safety and research company that builds the Claude family of AI models and apps, plus tools such as Claude Code for developers.
Who owns Anthropic?
Anthropic is a public benefit corporation owned by its stockholders. Amazon and Google are among its investors, and a Long-Term Benefit Trust helps elect its board.
Who founded Anthropic?
Anthropic was founded in 2021 by former members of OpenAI, including siblings Dario Amodei, the CEO, and Daniela Amodei, the president.
Is Anthropic the same as Claude?
No. Anthropic is the company, and Claude is its AI assistant and family of models.
Where is Anthropic based?
Anthropic is headquartered in San Francisco, California.
