The Stanford AI Index 2026 found that coding benchmark performance jumped from 60% to near 100% in a single year, enterprise adoption hit 88%, and China has erased the US lead in frontier model production.
Anthropic launched Claude Design, a prompt-to-prototype design tool that immediately sent Figma's stock nosediving — the latest in Anthropic's pattern of entering established software verticals with Claude-native products.
The UK government's AI Security Institute independently confirmed that Claude Mythos Preview can execute multi-stage cyberattacks autonomously — succeeding 73% of the time on expert-level tasks that no model could complete a year ago.