Anthropic is a US artificial intelligence company that builds the Claude family of large language models and publishes research on AI safety. It describes itself as a public benefit corporation, and it made its first public funding announcement, a USD 124 million Series A round, on May 28, 2021. This profile summarizes what the company and independent records state, with dates, as of October 23, 2026.
Key takeaways
- Anthropic states its purpose as the responsible development and maintenance of advanced AI for the long-term benefit of humanity.
- Its latest company-announced funding round that could be confirmed is a USD 30 billion Series G on February 12, 2026, at a USD 380 billion post-money valuation.
- Governance combines a board with a Long-Term Benefit Trust, a body of independent trustees created in 2023.
- Safety work includes the Constitutional AI method (2022) and a Responsible Scaling Policy, now at version 3.4.
Origins and the public benefit structure
The earliest company announcement in the public record is the May 28, 2021 Series A, led by technology investor Jaan Tallinn. In that post Anthropic called itself "an AI safety and research company" and named siblings Dario Amodei (CEO) and Daniela Amodei (President) as its leaders. The company's own pages do not give a founding date or a headquarters address, so neither is stated here.
Anthropic's company page says it is a public benefit corporation. Under the Delaware statute that governs many US companies of this type, such a corporation is profit-seeking but must be run in a responsible and sustainable manner, and its directors must balance stockholders' financial interests against the public benefit named in the charter and the interests of people materially affected by its conduct. Anthropic's stated purpose is "the responsible development and maintenance of advanced AI for the long-term benefit of humanity."
Funding rounds from company announcements
The table lists rounds that Anthropic announced on its own site. Figures are in US dollars and are the company's statements, not audited numbers.
| Date | Round | Amount raised | Post-money valuation |
|---|---|---|---|
| May 28, 2021 | Series A | USD 124 million | Not stated |
| March 3, 2025 | Series E | USD 3.5 billion | USD 61.5 billion |
| September 2, 2025 | Series F | USD 13 billion | USD 183 billion |
| February 12, 2026 | Series G | USD 30 billion | USD 380 billion |
The Series F announcement was led by ICONIQ, with Fidelity Management and Research Company and Lightspeed Venture Partners as co-leads, and said the company served over 300,000 business customers. The Series G was led by GIC and Coatue. News outlets have reported later and larger financing, but no company announcement of it could be confirmed, so it is left out of the table.
Governance: the board and the Long-Term Benefit Trust
As of October 2026 the company page lists six directors: Dario Amodei, Daniela Amodei, Yasmin Razavi, Reed Hastings, Chris Liddell, and Vas Narasimhan. Alongside the board sits the Long-Term Benefit Trust, which Anthropic described on September 19, 2023, as an independent body of financially disinterested members organized as a Delaware purpose trust. It holds a special class of stock that lets it elect and remove a share of the board, and Anthropic's October 2026 statement says the Trust has authority to appoint a majority of the board. The company page names four trustees: Ben Bernanke, Richard Fontaine, Neil Buddy Shah, and Christopher Stubbs, who joined on October 21, 2026.
On October 18, 2026, Anthropic announced an Anthropic Public Benefit Fund, to which it plans to commit 5% of its shares, subject to stockholder approval. The announcement says the fund's Anthropic equity will not be used for lobbying or political giving, and that further detail on structure and governance is still to come.
The Claude model family
Anthropic released Claude, with a lighter variant called Claude Instant, on March 14, 2023, through a chat interface and an API. In the company's news feed for autumn 2026, the Claude 5.5 generation includes Opus 5.5 (September 22), Sonnet 5.5 (September 28), and Haiku 5.5 (October 7), alongside Claude Mythos and Claude Fable 5.1 (September 15). Listed API prices per million tokens are USD 4 input and USD 20 output for Opus 5.5, and USD 2 input and USD 10 output for Sonnet 5.5. Prices and model lineups change often, so the company's pricing pages are the reference.
Safety research and policy documents
Three published items define most of the company's safety approach. The paper "Constitutional AI: Harmlessness from AI Feedback" (arXiv, December 15, 2022) describes training an assistant from a short list of written principles, with a model critiquing its own outputs and an AI system judging which of two responses is better. Anthropic published a plain-language constitution for Claude on May 9, 2023, and released a new version on January 21, 2026.
The Responsible Scaling Policy, first announced on September 19, 2023, ties safeguards to AI Safety Levels, a scale loosely modeled on biosafety levels. The policy page lists version 3.4, effective July 8, 2026, plus risk reports dated February and August 2026. These documents describe commitments by the company; independent verification of how they are applied is a separate question that the documents themselves do not answer.
What the public record does not settle
- Customer counts and revenue figures come from company statements, not audited filings.
- Valuations reflect private funding rounds and are not market prices.
- Trust and board membership changes frequently, so the company page should be checked for the current list.
- Comparisons with other AI developers depend on benchmark choice and date, and none are made here.
FAQ
What does Anthropic make?
The company develops the Claude models and offers them through a chat product, an API, and major cloud platforms including Amazon Web Services, Google Cloud, and Microsoft Azure, according to its Opus 5.5 announcement.
What is a public benefit corporation?
It is a for-profit corporation that names a public benefit in its charter. Under Delaware law, its board must weigh that benefit and the interests of affected people against stockholders' financial interests.
What is the Long-Term Benefit Trust?
It is an independent trust that holds a special class of Anthropic stock and has authority to elect and remove directors. Its trustees are chosen by the existing trustees and are not company employees or investors.
What is the Responsible Scaling Policy?
It is a published framework that links the safety, security, and deployment measures for a model to its assessed potential for catastrophic risk. The company revises it periodically and keeps a dated log of versions.
Related reading
- An early comparison of ChatGPT and Claude 2
- Profile of the language AI company Cohere
- How Grok from xAI works
- Overview of DeepSeek
More in AI Tools and Industry.
Last verified: October 2026
Sources
- Anthropic, Company page (governance, board, trustees), accessed October 2026
- Anthropic, Responsible Scaling Policy, version 3.4, 2026
- Anthropic, Anthropic raises $124 million to build more reliable, general AI systems, May 28, 2021
- Anthropic, Series E announcement, March 3, 2025
- Anthropic, Series F announcement, September 2, 2025
- Anthropic, Series G announcement, February 12, 2026
- Anthropic, The Long-Term Benefit Trust, September 19, 2023
- Anthropic, Christopher Stubbs joins the Long-Term Benefit Trust, October 21, 2026
- Anthropic, Investing in the public benefit, October 18, 2026
- Anthropic, Introducing Claude, March 14, 2023
- Anthropic, Claude Opus 5.5 and Claude Sonnet 5.5 announcements, September 2026
- Anthropic, Claude's Constitution, May 9, 2023, updated January 21, 2026
- Anthropic, Anthropic's Responsible Scaling Policy announcement, September 19, 2023
- Bai et al., Constitutional AI: Harmlessness from AI Feedback, arXiv:2212.08073, 2022
- State of Delaware, Delaware Code Title 8, Chapter 1, Subchapter XV (public benefit corporations)