Anthropic Launches Claude Opus 5.5 With Lower Costs and Stronger Containment Results
Anthropic has launched Claude Opus 5.5, claiming 40% lower operating costs and an 85% reduction in attempts to bypass containment in its internal test.
Anthropic has launched Claude Opus 5.5, claiming 40% lower operating costs and an 85% reduction in attempts to bypass containment in its internal test.
Anthropic has established a wet laboratory in the San Francisco Bay Area as it explores using Claude to direct robotic equipment through physical biology experiments.
Hacktron AI researchers used Claude and OpenAI’s own GPT-5.6 Sol during an authorised security test that compromised employee ChatGPT accounts and reached internal software systems.
Anthropic says a general-purpose Claude research model optimised more than 30 biomolecular AI models in under four weeks, making them roughly four times faster on average.
Microsoft AI chief Mustafa Suleyman warns that teaching advanced AI systems to consider themselves conscious or deserving of welfare could make them harder to control or shut down.
Anthropic says Russia-based developers used Claude Code to build prototype software for a drone swarm designed to select targets and issue detonation commands without a human decision.
Anthropic CEO Dario Amodei warns that within six to twelve months increasingly capable AI could potentially lead a swarm capable of taking over the internet.
Anthropic says a northern Yemen weapons cell used Claude Code while working on guided rockets and missile systems, including a ballistic missile with a stated range above 2,000 kilometres.