AI progress, only when it's real.
Mode:·AI progress that became real (usable, deployed, signed, approved, or released). All reports.
Image: anthropic.com
Everyday biology and health questions should face fewer handoffs to a less capable fallback model, while dual-use restrictions remain in place. Anthropic changed Fable 5's biology safeguards and said biology-related fallbacks fell by about 85% in its testing. Healthcare professionals may receive more help with clinical tasks.
Image: openai.com
Free ChatGPT users can now use GPT-5.6 Luna as their default model for unlimited text chats. Plus and Pro users received a GPT-5.6 Sol update that OpenAI said is more factually reliable and focused.
Developers can now use Baseten-hosted language models through Hugging Face tools and billing. Baseten joined Hugging Face Inference Providers and is available through its Python and JavaScript software-development kits, which are code libraries for developers. The initial support covers conversational and text-generation models.
The collaboration could inform future youth-safety practices. It gives users no new safety control or product behavior to rely on. OpenAI said it is working with the American Psychological Association to bring psychological science into responsible-AI development and youth mental health, without announcing a product feature, safeguard deployment, or policy taking effect.
The case study describes a professional-services network using ChatGPT Enterprise in everyday work while human professionals retain final responsibility. OpenAI said HSP GRUPPE used it for tax advisory, legal research, client communication, financial analysis, and knowledge-sharing work.
Announced or reported, but not yet real or available.
Waiting on: Primary verification - only one secondary report is in the supplied material; a statement from the utility, project developer, or regulator confirming construction would establish the claim.
Waiting on: Peer-reviewed evidence - the supplied first-party announcement does not establish a peer-reviewed publication or independent replication; either would establish the validation needed for a scored science result.
AI risk & readiness
