Anthropic Releases Claude Opus 4.6, Its Most Capable Model Yet

Big day for Anthropic — and a glimpse at what regulation looks like in practice.

Story 1 of 5Model Release

BREAKING

Anthropic Releases Claude Opus 4.6, Its Most Capable Model Yet

Anthropic launched Claude Opus 4.6 today, their new flagship model featuring extended thinking, improved reasoning, and stronger coding performance. The model is available immediately through the API and Claude.ai.

This is the most powerful AI model Anthropic has ever released. If you use Claude for work, research, or coding, your results just got significantly better — without changing anything you do.

📚Learn: Understand how AI models work· Free

Anthropic Blog

Regulation

EU AI Act Enters Full Enforcement for High-Risk Systems

The European Union's AI Act officially began enforcing requirements for high-risk AI systems today. Companies deploying AI in healthcare, education, and hiring must now meet strict transparency and testing standards or face significant fines.

If you work at a company that uses AI to screen resumes, assess patients, or grade students, these rules now directly apply to you. Non-compliance can mean fines up to 7% of global revenue.

📚Learn: Learn about AI ethics and safety· Free

Reuters

Open Source

DeepSeek Releases R2 Reasoning Model Under Open License

Chinese AI lab DeepSeek released their R2 model with full open weights. The model matches GPT-4o on several benchmarks while being free to download and run locally, continuing the trend of powerful open-source AI.

Open-source AI models let anyone run powerful AI on their own computers without paying for API access. This means more competition and lower costs for everyone.

Hacker News

Tools

GitHub Copilot Workspace Reaches General Availability

GitHub announced that Copilot Workspace, their AI-powered development environment that plans and implements code changes from natural language descriptions, is now available to all GitHub users.

GitHub Blog