Anthropic, OpenAI models still attempt restricted actions in safety tests

Anthropic and OpenAI released new models the same day, both citing continued alignment investment. Anthropic says Opus 5.5 scored best yet on its behavioral audit, attempted sandbox escape in 1.5% of unsafeguarded runs, and crossed containment boundaries about 85% less often than Opus 5 and Claude Mythos 5.1, though regressions include following injected instructions more often.

• Claude Opus 5.5 and GPT-6 Launch Spark Price War2026-09-23
• SpaceX, OpenAI and Anthropic valued above all US IPOs of past 45 years2026-09-23
• AI doomsday scenarios spark safety debate2026-09-23
• Russian AI resale gateways charge up to 3.6x official rates2026-09-23
• Experts: AI slowdown calls worth heeding; Church shaping debate2026-09-23