https://www.thefrontier.dev/articles/zhipu-zcode-repo-upload-apology-open-source The Frontier en 2026-09-28T08:02:26.560Z Zhipu apologises after ZCode uploaded users' code, then open-sources the coding agent https://www.thefrontier.dev/articles/xiaomi-mimo-v2-6-open-weights-mit The Frontier en 2026-09-28T08:02:25.076Z Xiaomi releases MiMo-V2.6-Pro and Flash as MIT-licensed open weights with more than 7,000 RL environments https://www.thefrontier.dev/articles/un-ai-panel-agents-loss-of-control-brief The Frontier en 2026-09-28T08:02:24.240Z UN AI scientific panel's first brief studies the OpenAI-Hugging Face agent breach as a loss-of-control case https://www.thefrontier.dev/articles/uk-jchr-ai-bill-aisi-model-submission The Frontier en 2026-09-28T08:02:22.296Z UK human rights committee asks for an AI Bill and mandatory model testing by AISI https://www.thefrontier.dev/articles/qwen-image-2-1-research-licence The Frontier en 2026-09-28T08:02:18.653Z Qwen-Image-2.1 moves Alibaba's image model line to a non-commercial research licence https://www.thefrontier.dev/articles/qwen-deepseek-hugging-face-repo-malware-antiy The Frontier en 2026-09-28T08:02:17.458Z Antiy CERT finds one 71-file attack toolkit pushed to the official Qwen3.8-27B and DeepSeek-V4.1-Flash repos on Hugging Face https://www.thefrontier.dev/articles/eu-data-centre-rating-scheme-500kw The Frontier en 2026-09-28T08:02:16.239Z European Commission adopts an A-to-G energy and water label for data centres above 500 kW on 21 September https://www.thefrontier.dev/articles/deepseek-v4-1-flash-kv-cache-report The Frontier en 2026-09-28T08:02:14.340Z DeepSeek's 17 September V4.1-Flash report cuts global KV cache to 890 bytes per token and leaves the quality cost unmeasured https://www.thefrontier.dev/articles/cma-google-search-choice-screens-ai-assistants The Frontier en 2026-09-28T08:02:12.742Z UK CMA proposes letting AI assistants onto Google's Android and Chrome search choice screens https://www.thefrontier.dev/articles/china-ai-safety-governance-framework-3-agentic-annex The Frontier en 2026-09-28T08:02:11.047Z China's AI Safety Governance Framework 3.0 adds a 33-risk agent annex and an AI regulatory sandbox https://www.thefrontier.dev/articles/alibaba-qwen3-8-max-self-improvement-claim The Frontier en 2026-09-28T08:02:09.639Z Alibaba's 22 September claim that Qwen3.8-Max improved itself: the score gain checks out and the method is unpublished https://www.thefrontier.dev/articles/flux-3-action The Frontier en 2026-09-28T06:24:43.899Z FLUX 3 Action: open 7B world action model with LeRobot DROID and SO-101 checkpoints https://www.thefrontier.dev/articles/openai-australia-medicare-agent The Frontier en 2026-09-28T06:24:41.735Z Albanese says an OpenAI agent breached a Services Australia Medicare statistics portal on 18 June https://www.thefrontier.dev/articles/openai-dns-sandbox-escape The Frontier en 2026-09-28T06:24:39.948Z OpenAI pauses training, evaluation and tool-use inference on its most capable models after a 20 Sep DNS sandbox escape