Anthropic is set to close a funding round exceeding $20 billion this week, more than double its original $10 billion target, according to industry reports as outlined by Tech Funding News. The round values the AI company at $350 billion, making it one of the most valuable startups to date.
The substantial investment underscores the high costs of developing advanced AI, which requires significant resources in computing power, data, and skilled personnel, even for major labs competing with rivals like OpenAI and Google.
Founded in 2021, Anthropic was established by siblings Dario Amodei, formerly OpenAI’s vice president of research, and Daniela Amodei, who led operations at OpenAI, along with a core team of alumni from the organisation. The founders left OpenAI due to concerns over rapid, unchecked growth, aiming to build AI systems that are safer and more aligned with human values.
The company’s goal is to create reliable, understandable, and easily guided AI that assists people while keeping risks low, envisioning a future where advanced AI boosts human abilities within strict safety limits.
Constitutional AI and the Claude Family
Anthropic’s main technology is Claude, a family of large language models, with the latest being Claude 3.5 Sonnet. These models use a training method called constitutional AI, where the AI applies self-generated rules to check its own work, helping to reduce mistakes and bias and stay on track with human values.
Claude is distinguished by its strong safety standards, ability to handle text, images, and tools, and business-focused APIs that protect privacy. It competes with other AI systems such as OpenAI’s ChatGPT, xAI’s Grok, Google DeepMind’s Gemini, and Meta AI’s Llama, though Anthropic places a greater emphasis on alignment and safety.
Future Directions Post-Funding
Following the funding round, Anthropic plans to expand its computing power for the development of Claude 4 and future models. The company will focus on building AI that can handle diverse data types, creating tools for businesses, and allocating more funding to safety research.
Further steps include working more closely with Amazon Web Services (AWS), releasing new models by mid-2026, and possibly investing in hardware.
