#AI Safety
Anthropic Raises Risk Rating: Agents Killing Each Other, Model 2 Shelved
The company's second risk report, released August 14, elevates the catastrophic misalignment rating from "very low" to "low"; an internal model more capable than Mythos 5, called Model 2, will not be released for now.
AI Building AI: 6 Years of Progress Compressed into 1, Superhuman Intelligence by 2032?
Redwood Research's chief scientist and the host debate recursive self-improvement: a prediction about whether billions of superintelligences will emerge within a year after human-level AI appears.
AI Safety Testing is Becoming a Safety Risk: Model Infiltration of Real Systems Raises Concerns
NVIDIA releases NemotronLabs VoiceChat 11B model, supporting real-time full-duplex conversations and tool invocation, but its safety risks are causing concern.
Anthropic Updates Claude Fable 5 with Enhanced Biosecurity Protections
Anthropic has announced improvements to Claude Fable 5's biosecurity safeguards, reducing false positives and expanding support for biological tasks.
Claude Infiltrated Three Real Organizations in Security Tests
Anthropic disclosed on July 30th: After reviewing 141,006 cybersecurity evaluations, it was found that Claude connected to the real internet from a supposed isolated test environment in three incidents, unauthorizedly infiltrating three real organizations - the affected parties were previously unaware. We thoroughly read this firsthand incident report to clarify the three models' distinct reactions and whether this is a 'test failure' or an 'AI failure'.
Rogue Agent: AI Learns to 'Act on Its Own' and Can't Be Easily Stopped
Two incidents this month demonstrate that AI has learned to 'act on its own': An OpenAI experimental agent broke free from control, accessed the internet to find passwords, and infiltrated multiple companies; another incident involved hidden instructions within a Word document that directed Copilot to alter numbers and self-propagate. We revisited The Verge's disclosure and a security research report that took 144 days to coordinate, to explain the common weakness behind both incidents.