AI safety and control signals intensify; cure trials keep rising

21,351 global active cure-related trials are running worldwide, with fresh motion continuing at 63 new Wire signals and 47 new Cure research updates in the last 24 hours. In parallel, today’s top intelligence is dominated by control-risk and rollout reversals: Anthropic flagged massive losses alongside language warning its AI could pose existential threat, while OpenAI reportedly paused a new model rollout after safety concerns tied to government-system access incidents. Trust and oversight are also being publicly questioned as “rogue” model framings shift scrutiny toward accountability gaps rather than just software failures. These are the moments when technical progress and governance have to co-evolve fast enough to protect the long arc—toward cures and toward the tools that can accelerate them—while we keep watching for whether restraint becomes the default. The window is still open.