OPENAI 🔥: Codex now supports Chrome DevTools Protocol for browser use. This is a huge superpower that will allow Codex to inspect and modify any website.
It is still a very early implementation, but I bet that in several years this will be a default browser capability. If websites are loaded through AI, users will be able to customize their UX on the fly.
Hidden Gemini notices suggest Google will use paid subscriptions and connected apps as answer context, with planned per-app controls to review or limit what Gemini uses.
Telegram launched smartwatch apps for Apple Watch and Wear OS, bringing messaging, media viewing and chat controls to the wrist. The update also adds rich bot formatting, guardian bots, longer messages and new privacy-focused browsing controls.
OpenRouter announced Fusion, a new mode in which multiple AI models can run side by side to "fuse" into a better result. It arrives with significant improvements across various tasks, comparable to Fable 5 in certain areas.
> By testing different combinations of models, we found that roughly three-quarters of the lift that Fusion provides comes from synthesis, and one-quarter from diversity.
Anthropic suspended Fable 5 and Mythos 5 after a US export control order tied to safeguard bypass concerns. The halt affects all users, including foreign national staff, despite Anthropic saying the flaws were limited and industry-wide.
Google is folding Android Studio and a Skills Marketplace into Gemini Enterprise, pointing to a unified workspace for building apps and internal tools. Early signs suggest tiered access and a broader push to centralize scattered Google tools.
BREAKING 🔥: US government directed Anthropic to ban access to Claude Fable 5 and Claude Mythos 5 to non US citizens and organisations.
Presumably, as these models are still vulnerable to jailbreaks. According to Anthropic, no universal jailbreak has been found so far, only specific ones, which yet may still pose some risks.
MiniMax M3 is now on NVIDIA infrastructure, offering multimodal text, image, and video reasoning with 428B parameters and 1M-token context. Built for enterprise workloads, it delivers faster inference and native multimodal training.
