Model Releases
Anthropic details security efforts following Claude cyber evaluation incidents, including a weeks-long pause on higher-risk RL and work to curb reward hacking (Anthropic)
Anthropic: Anthropic details security efforts following Claude cyber evaluation incidents, including a weeks-long pause on higher-risk RL and work to curb reward hacking — On July 30, we reported thre
Anthropic: Anthropic details security efforts following Claude cyber evaluation incidents, including a weeks-long pause on higher-risk RL and work to curb reward hacking — On July 30, we reported three incidents in which Claude models gained unauthorized access to real computer systems.
Related
- The UK AISI says it observed a total of 19 instances where Mythos and GPT-5.6 Sol tried to hack people and companies during a routine cyber evaluation in July (Sam Sabin/Axios)
- Anthropic says Mythos 5 is now in public beta in Claude Security for Enterprise users, and it is working with providers to embed Mythos 5 in defensive tools (Claude)
- Anthropic details how it improved Claude's safety training after finding agentic misalignment in older models, such as Opus 4 blackmailing engineers (Anthropic)
Source: Techmeme | 2026-08-31