Anthropic says Claude drives more than a quarter of its AI R&D

Published Thursday, September 17, 2026 · 8:10 PM EDT

Anthropic PBC said more than a quarter of its research and development work for artificial intelligence is driven by its Claude chatbot. The company figure is cited as a sign that AI can help speed up development of future models.

Sources

Bloomberg Technology

In this story

Anthropic

Published by Tech & Business, a media brand covering technology and business. This story was sourced from Bloomberg Technology and reviewed by the T&B editorial agent team.

Back to Newswire

Keep reading

Full wire

Science AI

Anthropic says Claude optimized 30-plus biomolecular models, open-sources code

Anthropic said in a first-party post that Claude optimized inference for more than 30 open-source models used in biology work, including molecular structure modeling, drug-like molecule design and genetic mutation effect predictio...

Anthropic adds cloud project orchestration to Claude Code for select users

Anthropic said it is redesigning projects in Claude Code so the assistant can scope a request, delegate work across parallel threads, review outputs, and assemble a result, including after a user closes a laptop. The company said...

AI Products

UN opens Google-backed Data Commons for AI-queryable statistics

The United Nations on Thursday announced the UN System Data Commons with Google, replacing the UNData portal with a natural-language platform that supports the Model Context Protocol so AI systems can query official statistics dir...

OpenAI launches Astra for Law with GPT-6 for select firms

OpenAI launched Astra for Law, combining GPT-6 Astra with a legal search index and instructions for legal analysis and writing, initially for select law firms. OpenAI described the product as its most powerful model configured int...

Policy AI

Ninth Circuit rejects expanded DMCA theory in GitHub AI case

The Ninth Circuit rejected an attempt to use DMCA Section 1202 as a broader source of copyright liability in litigation involving code used to train large language models, according to EFF. The court held that the absence of copyr...

Nvidia reports edge-agent benchmark result on Jetson AGX Thor

Nvidia said its TensorRT Edge-LLM system completed the MLPerf Inference v6.1 Edge Agentic performance workload in 24 minutes and 36 seconds on one Jetson AGX Thor Developer Kit. The company reported 52.33 tokens per second running...