Claude Now Leads 26% of Anthropic’s AI R&D Work
Anthropic says Claude now leads 26% of its measured AI R&D work. What does this mean for AI development, oversight and human-in-the-loop research?
Tag: Frontier AI
DeepSeek V4.1-Flash Launches: Why Faster, Smaller AI Models Could Reshape the AI Race
DeepSeek launched V4.1-Flash on September 10, 2026, describing it as the smallest model in a new architecture designed for faster inference, higher throughput and scaling. The release shows why AI economics is shifting from sheer model size toward useful intelligence per unit of compute.
OpenAI Says Astra Needs Stronger Guardrails: AI Capability Is Beginning to Trigger Its Own Safety Thresholds
OpenAI says its forthcoming Astra model is the first to trigger stronger safeguards under its safety protocol because of advanced cybersecurity capabilities. The milestone raises a bigger question: what happens when AI capability itself determines how tightly a model must be controlled?
OpenAI Puts the Brakes on Astra After Tests Suggest It May Reach “Critical” Cyber Capability
OpenAI has tightened controls around Astra after preliminary tests suggested the unreleased model may reach its highest cybersecurity risk threshold.
Sam Altman Says AI Development May Need to Slow Down—and Nobody Wants to Brake First
Sam Altman says frontier AI may need to slow temporarily so society can build safeguards, but competitive pressure makes unilateral action unlikely.
Britain Says It May Regulate Frontier AI if Voluntary Safety Agreements Fail
The UK says it may introduce stronger laws for advanced AI models if voluntary safety testing fails to protect the public from emerging risks.