This site demonstrates one possible use of this domain. For acquisition, partnership, or investment inquiries, please use our contact link. (brainmatter.com)
BRAINMATTER - Intelligence Beyond Limits

Anthropic

Safety-first frontier lab building Claude, with a Constitutional AI training methodology.

Official site Lab directory entry

History

Founded in 2021 by Dario and Daniela Amodei and several senior OpenAI researchers who left over disagreements about commercialisation pace and safety prioritisation. Established as a Public Benefit Corporation.

Founders & leadership

Founders

  • Dario Amodei
  • Daniela Amodei
  • Tom Brown
  • Sam McCandlish
  • Jack Clark
  • Jared Kaplan

Current leadership

  • Dario Amodei - CEO
  • Daniela Amodei - President
  • Jared Kaplan - Chief Science Officer

Funding

Over $15B raised; major investors include Google ($2B+) and Amazon ($4B+).

Major models & products

  • Claude 1–3

    Progressive frontier models; Claude 3 Opus matched GPT-4 in 2024.

  • Claude 3.5 / 3.7 Sonnet

    Best-in-class for coding and tool use through 2024–2025.

  • Claude Opus 4

    2025 flagship model.

Research programs

  • Frontier model research
  • Mechanistic interpretability
  • Responsible Scaling Policy

Safety initiatives

  • Constitutional AI training (CAI)
  • Responsible Scaling Policy (RSP) with AI Safety Level commitments
  • Mechanistic interpretability team publishing extensively

Major publications

Key breakthroughs

  • Constitutional AI 2022

    Replaced parts of RLHF with AI-generated critiques against a written constitution.

  • Computer Use (Claude) 2024

    First major lab to ship general-purpose desktop agency.

Partnerships

  • Amazon AWS (primary cloud)
  • Google Cloud
  • Palantir
  • US AI Safety Institute

Timeline

  1. 2021

    Founded by ex-OpenAI researchers.

  2. 2022

    Claude alpha; Constitutional AI paper.

  3. 2023

    Amazon and Google investments.

  4. 2024

    Claude 3 family; Computer Use.

  5. 2025

    Claude Opus 4.

Competitive landscape

Most directly comparable to OpenAI. Differentiated by safety-research emphasis, the Constitutional AI methodology, and a stronger enterprise/government posture than consumer focus.

Other intelligence centers

OpenAI

Capped-profit frontier lab that defined the generative AI era with the GPT series.

Google DeepMind

Google's unified AI research arm - the home of AlphaGo, AlphaFold, and Gemini.

Meta AI

Meta's open-weights frontier program - the home of Llama and FAIR.

xAI

Elon Musk's frontier lab, builder of Grok and the Colossus supercluster.

NVIDIA Research

The hardware platform underpinning the entire modern AI industry - and a serious research lab in its own right.

Mistral AI

European open-weights frontier lab founded by ex-Meta and DeepMind researchers.