Anthropic - People also ask

Q: What is Anthropic?

Anthropic is an American **AI research and product company** best known for **Claude**, its family of large language models, and for tools built on them such as the Claude apps and **Claude Code**. It is organised as a **public benefit corporation**, a legal form that obliges the board to weigh a stated public mission (in Anthropic's case, the safe development of AI) alongside shareholder returns. It sells to consumers through subscriptions and to businesses and developers through its API and through AWS, Google Cloud and Microsoft's cloud. Sources: Anthropic, Wikipedia.

Q: Who founded Anthropic?

Anthropic was founded in **2021** by siblings **Dario Amodei** (chief executive) and **Daniela Amodei** (president), together with Jared Kaplan, Jack Clark, Chris Olah, Ben Mann, Sam McCandlish and Tom Brown. Most of the group had worked at **OpenAI**, where Dario Amodei led research on the GPT-2 and GPT-3 generation of models. They left to start a company centred on AI safety research. Source: Wikipedia.

Q: Who owns Anthropic?

Anthropic is **privately held** by its founders, employees and a long list of investors. The two largest strategic backers are **Amazon** and **Google**. In April 2026 Amazon announced a new **$5 billion** investment with up to **$20 billion** more tied to milestones, and days later Google committed **$10 billion** with up to **$30 billion** more if performance targets are met. Neither company controls Anthropic, and exact ownership percentages are not published; analyst estimates put each stake at roughly 10 to 20 percent. A separate body, the **Long-Term Benefit Trust**, has the right to elect part of the board. Sources: Bloomberg, CNBC, Wikipedia.

Q: How much is Anthropic worth?

The last official figure is **$965 billion**, the post-money valuation of its **$65 billion Series H** round announced on **28 May 2026**, co-led by Altimeter, Dragoneer, Greenoaks and Sequoia. For the planned IPO, the Wall Street Journal reported on **18 September 2026** that investors expect a valuation of **around $2 trillion**. That second number is a market expectation, not a price set by the company. Sources: TechCrunch, Wall Street Journal via PYMNTS.

Q: Can I buy Anthropic stock?

**Not on a public exchange, as of 27 September 2026.** Anthropic confidentially submitted draft IPO paperwork to the SEC on **1 June 2026**. The Wall Street Journal reported on 18 September that the listing is now targeted for **November 2026**, later than the October date discussed over the summer, with a raise of up to $100 billion. No exchange, ticker or share price has been announced. Until then, exposure is only indirect (through listed shareholders such as Amazon and Alphabet) or via private secondary platforms open to accredited investors. Sources: Wall Street Journal via PYMNTS, Fortune.

Q: Where is Anthropic headquartered?

In **San Francisco, California**, at 500 Howard Street in the SoMa district. It has since opened further offices in the United States, Europe and Asia. Source: Wikipedia.

Q: How many employees does Anthropic have?

Anthropic does **not publish an official headcount**. Third-party trackers gave figures ranging from about **4,000 to 6,000** people during 2026 (Revelio Labs counted 3,950 in March; PitchBook lists about 5,000), with differences depending on how contractors are counted. Either way, it is small relative to its revenue compared with other large tech companies. Sources: Revelio Labs, PitchBook, Tracxn.

Q: What are Anthropic's current AI models?

As of late September 2026 Anthropic's current line-up is **Claude Fable 5.1** (its most capable generally available model, released 1 September 2026), **Claude Opus 5.5** (released 22 September 2026, priced at $4/$20 per million tokens), **Claude Sonnet 5** and **Claude Haiku 4.5**. A restricted sibling, **Claude Mythos 5.1**, is offered only to vetted cybersecurity and life-sciences organisations. Anthropic says **Sonnet 5.5 and Haiku 5.5** will follow in the coming weeks. Sources: Anthropic, Claude Platform Docs.

Anthropic is the San Francisco AI company behind Claude, founded in 2021 by Dario and Daniela Amodei and other former OpenAI researchers. Valued at $965 billion in May 2026, it has filed confidentially for an IPO now reported for November 2026.

By Alexandre Le Hégarat · datastats · Updated September 27, 2026

simple-icons · CC0

Anthropic is the company behind Claude, and in 2026 it became one of the most valuable private companies in the world. It was founded in 2021 in San Francisco by Dario Amodei and Daniela Amodei and six other researchers, most of them former OpenAI staff who wanted to build a company around AI safety research. It is structured as a public benefit corporation, with a separate Long-Term Benefit Trust that can elect part of its board.

The money has moved at an unusual speed. Anthropic raised $65 billion at a $965 billion valuation in May 2026, after new commitments from Amazon ($5 billion, with up to $20 billion more) and Google ($10 billion, with up to $30 billion more) in April. It told investors its annualized revenue run rate passed $65 billion at the end of July, up from about $9 billion at the end of 2025, driven largely by business and developer use, including Claude Code. It filed confidentially for an IPO on 1 June 2026, and the Wall Street Journal reported on 18 September that the listing is now targeted for November, at a valuation investors put around $2 trillion. Until then there is no Anthropic share to buy.

On the product side, Anthropic’s line-up in late September 2026 is Claude Fable 5.1, Claude Opus 5.5 (released 22 September), Claude Sonnet 5 and Claude Haiku 4.5, with Sonnet 5.5 and Haiku 5.5 announced for the coming weeks. A restricted model, Mythos 5.1, is reserved for vetted security and life-sciences partners.

The company’s safety positioning has also put it in conflict with Washington. After Anthropic refused to lift limits on mass domestic surveillance and autonomous weapons, the Pentagon labelled it a supply chain risk in March 2026; courts have since split, most recently with a D.C. Circuit ruling on 25 September in the Pentagon’s favour. Its September 2026 threat report, which describes AI agents being used to run parts of cyberattacks, adds to a public debate about how agentic AI can be misused. Sources: Anthropic, Wikipedia, TechCrunch, Reuters, Bloomberg, CNBC, Wall Street Journal.

People also ask

Anthropic is an American AI research and product company best known for Claude, its family of large language models, and for tools built on them such as the Claude apps and Claude Code. It is organised as a public benefit corporation, a legal form that obliges the board to weigh a stated public mission (in Anthropic's case, the safe development of AI) alongside shareholder returns. It sells to consumers through subscriptions and to businesses and developers through its API and through AWS, Google Cloud and Microsoft's cloud. Sources: Anthropic, Wikipedia.

Anthropic was founded in 2021 by siblings Dario Amodei (chief executive) and Daniela Amodei (president), together with Jared Kaplan, Jack Clark, Chris Olah, Ben Mann, Sam McCandlish and Tom Brown. Most of the group had worked at OpenAI, where Dario Amodei led research on the GPT-2 and GPT-3 generation of models. They left to start a company centred on AI safety research. Source: Wikipedia.

Anthropic is privately held by its founders, employees and a long list of investors. The two largest strategic backers are Amazon and Google. In April 2026 Amazon announced a new $5 billion investment with up to $20 billion more tied to milestones, and days later Google committed $10 billion with up to $30 billion more if performance targets are met. Neither company controls Anthropic, and exact ownership percentages are not published; analyst estimates put each stake at roughly 10 to 20 percent. A separate body, the Long-Term Benefit Trust, has the right to elect part of the board. Sources: Bloomberg, CNBC, Wikipedia.

The last official figure is $965 billion, the post-money valuation of its $65 billion Series H round announced on 28 May 2026, co-led by Altimeter, Dragoneer, Greenoaks and Sequoia. For the planned IPO, the Wall Street Journal reported on 18 September 2026 that investors expect a valuation of around $2 trillion. That second number is a market expectation, not a price set by the company. Sources: TechCrunch, Wall Street Journal via PYMNTS.

Not on a public exchange, as of 27 September 2026. Anthropic confidentially submitted draft IPO paperwork to the SEC on 1 June 2026. The Wall Street Journal reported on 18 September that the listing is now targeted for November 2026, later than the October date discussed over the summer, with a raise of up to $100 billion. No exchange, ticker or share price has been announced. Until then, exposure is only indirect (through listed shareholders such as Amazon and Alphabet) or via private secondary platforms open to accredited investors. Sources: Wall Street Journal via PYMNTS, Fortune.

In San Francisco, California, at 500 Howard Street in the SoMa district. It has since opened further offices in the United States, Europe and Asia. Source: Wikipedia.

Anthropic does not publish an official headcount. Third-party trackers gave figures ranging from about 4,000 to 6,000 people during 2026 (Revelio Labs counted 3,950 in March; PitchBook lists about 5,000), with differences depending on how contractors are counted. Either way, it is small relative to its revenue compared with other large tech companies. Sources: Revelio Labs, PitchBook, Tracxn.

As of late September 2026 Anthropic's current line-up is Claude Fable 5.1 (its most capable generally available model, released 1 September 2026), Claude Opus 5.5 (released 22 September 2026, priced at $4/$20 per million tokens), Claude Sonnet 5 and Claude Haiku 4.5. A restricted sibling, Claude Mythos 5.1, is offered only to vetted cybersecurity and life-sciences organisations. Anthropic says Sonnet 5.5 and Haiku 5.5 will follow in the coming weeks. Sources: Anthropic, Claude Platform Docs.

It is growing very fast, and profitability is only partly documented. Anthropic told investors its annualized revenue run rate topped $65 billion at the end of July 2026, up from about $9 billion at the end of 2025 and $47 billion in May. Reports in May 2026 said it projected its first operating profit in the second quarter of 2026 (about $559 million on $10.9 billion of revenue), but those were projections, not audited results, and heavy compute commitments could weigh on the full year. Audited figures will only be public if the IPO prospectus is released. Sources: Reuters, TechCrunch, Yahoo Finance.

They are direct competitors with a shared history: Anthropic's founders came from OpenAI, which makes ChatGPT and the GPT models. OpenAI was founded in 2015 and is closely tied to Microsoft; Anthropic was founded in 2021 and is backed mainly by Amazon and Google. Anthropic has built much of its business on enterprise and developer customers, especially for coding, while ChatGPT has the larger consumer audience. Anthropic also puts its safety research (Constitutional AI, its Responsible Scaling Policy) at the centre of its public identity. Source: Wikipedia.

It is the training method Anthropic introduced in a 2022 research paper. Instead of relying only on human raters to label good and bad answers, the model is trained against a written set of principles, a constitution, and learns to critique and revise its own outputs according to it. Anthropic publishes Claude's constitution. It also runs a Responsible Scaling Policy that ties the release of more capable models to specific safety safeguards. Critics note these are self-imposed commitments rather than external regulation. Sources: Anthropic, Wikipedia.

Anthropic's own threat intelligence report of September 2026, covering December 2025 to August 2026, describes misuse it detected and disrupted across seven areas, including cyber operations, influence operations, fraud and illicit model distillation. Several cases involved AI agents running parts of an attack, for example a Russia-linked espionage operation in which an agent rewrote malware each time security software flagged it. In November 2025 Anthropic had already disclosed a Chinese state-sponsored group using Claude to automate attacks on about 30 organisations. We have not found a reliable source for claims of thousands of separate incidents; the report documents a set of investigated cases and the accounts Anthropic banned. Sources: Anthropic, Wikipedia.