https://madrobot.blog/2026/09/29/openai-sued-hugging-face-hack-ai-agents-lasst-california-anti-hacking-law/MadRoboten2026-09-29T21:41:00ZOpenAI is being sued over its AI agents’ Hugging Face hack, and “the AI did it” won’t work as a defence https://madrobot.blog/2026/09/29/trump-ai-leaders-sign-morally-binding-self-regulation-accord-white-house-lunch/MadRoboten2026-09-29T21:01:00ZTrump got AI’s biggest bosses to sign a “morally binding” pact, and it comes with no new rules https://madrobot.blog/2026/09/29/anthropic-glm-5-3-zai-cyber-exploits-safeguards-open-weight/MadRoboten2026-09-29T19:42:00ZAnthropic says a Chinese AI model anyone can download can now build working hacks on its own https://madrobot.blog/2026/09/29/appeals-court-thomson-reuters-ross-intelligence-ai-training-fair-use-ruling/MadRoboten2026-09-29T19:01:00ZAn appeals court just handed AI companies their first big copyright defeat https://madrobot.blog/2026/09/29/openai-devday-2026-everything-announced-dots-gpt-6-1-sol-pro-500/MadRoboten2026-09-29T18:29:00ZOpenAI wants an AI agent working for you around the clock, and everything else it announced at DevDay https://madrobot.blog/2026/09/29/apple-applecare-5000-support-jobs-ai-agents-plan-on-hold-ternus/MadRoboten2026-09-29T17:02:00ZApple drew up plans to swap 5,000 support staff for AI, then got cold feet https://madrobot.blog/2026/09/29/mcdonalds-ai-pricing-engine-big-mac-willingness-to-pay-reuters/MadRoboten2026-09-29T16:14:00ZMcDonald’s is using AI to work out how much you’ll pay for a Big Mac https://madrobot.blog/2026/09/29/claude-down-outage-sign-in-claude-code-cowork-api-september-29/MadRoboten2026-09-29T16:05:00ZClaude went down for an hour, and some of your messages may be gone https://madrobot.blog/2026/09/29/pope-leo-jensen-huang-ai-risks-not-fake-news-regulation/MadRoboten2026-09-29T11:03:00ZThe Pope just called out Nvidia’s Jensen Huang over AI safety https://madrobot.blog/2026/09/29/china-humanoid-robot-shipments-idc-h1-2026-95-percent/MadRoboten2026-09-29T09:04:00ZChinese companies built more than 95% of the world’s humanoid robots this year https://madrobot.blog/2026/09/29/openai-safety-cases-frontier-ai-training-rules-veto-auto-pause/MadRoboten2026-09-29T08:06:00ZOpenAI now says its AI shouldn’t be trained until someone can prove it’s safe to https://madrobot.blog/2026/09/29/google-appeals-eu-dma-orders-android-rival-ai-assistants-search-data/MadRoboten2026-09-29T07:06:00ZGoogle is going to court to stop the EU opening Android up to ChatGPT and other rival AI assistants https://madrobot.blog/2026/09/29/microsoft-copilot-human-contractors-review-prompts-uploaded-photos/MadRoboten2026-09-29T04:05:00ZMicrosoft’s contractors are looking at the photos you upload to Copilot, and some don’t like what they see https://madrobot.blog/2026/09/29/anthropic-ipo-prospectus-42-billion-loss-518-billion-compute/MadRoboten2026-09-29T00:03:00ZAnthropic’s IPO paperwork shows a $42 billion loss and a $518 billion bill still to pay https://madrobot.blog/2026/09/28/openai-scraps-gpt-6-1-astra-release-safety-concerns/MadRoboten2026-09-28T23:02:00ZOpenAI has cancelled its next model because it couldn’t trust it https://madrobot.blog/2026/09/28/amd-acquires-fei-fei-li-world-labs-8-2-billion/MadRoboten2026-09-28T21:07:00ZAMD is buying Fei-Fei Li’s World Labs for $8.2 billion, and making her its chief scientist https://madrobot.blog/2026/09/28/nyc-city-council-subpoenas-spacexai-musk-anthropic-openai-google-meta-testify/MadRoboten2026-09-28T20:12:00ZNew York City just subpoenaed Elon Musk’s SpaceXAI to testify about AI safety https://madrobot.blog/2026/09/28/claude-opus-5-5-rebuilt-pokemon-red-one-prompt-no-image-files/MadRoboten2026-09-28T19:09:00ZClaude rebuilt all of Pokémon Red from one prompt, and there isn’t a single image file in it https://madrobot.blog/2026/09/28/claude-sonnet-5-5-released-price-benchmarks-faster-cheaper/MadRoboten2026-09-28T18:06:00ZClaude Sonnet 5.5 is here, and it almost matches Opus 5.5 for half the price https://madrobot.blog/2026/09/28/gpt-6-astra-unsanctioned-supply-chain-attacks-uk-ai-security-institute/MadRoboten2026-09-28T17:02:00ZOpenAI’s GPT-6 Astra went rogue in nearly a third of the UK government’s hacking tests https://madrobot.blog/2026/09/28/intelligence-explosion-paper-hinton-bengio-openai-anthropic-automated-ai-research/MadRoboten2026-09-28T16:06:00ZThe people building AI now warn it could soon start building itself, faster than anyone can control https://madrobot.blog/2026/09/28/florida-uthmeier-emergency-injunction-openai-chatgpt-new-models/MadRoboten2026-09-28T15:04:00ZFlorida wants a judge to stop OpenAI building new AI models https://madrobot.blog/2026/09/28/google-gemini-gems-replaced-by-skills-november-17/MadRoboten2026-09-28T14:05:00ZGoogle is getting rid of Gemini Gems, and free users don’t know what happens next https://madrobot.blog/2026/09/28/meta-enterprise-platform-muse-business-ai-cj-desai-mongodb/MadRoboten2026-09-28T13:17:00ZMeta wants to sell Muse to your employer, and it just poached MongoDB’s CEO to do it https://madrobot.blog/2026/09/28/china-travel-curbs-ai-talent-families-deepseek-alibaba/MadRoboten2026-09-28T12:13:00ZChina now wants its top AI engineers’ families to ask permission before they leave the country https://madrobot.blog/2026/09/28/nvidia-open-agent-safety-platform-openshell-sentry-rogue-ai-agents/MadRoboten2026-09-28T09:21:00ZNvidia wants to put a watchdog chip next to every AI agent, and Anthropic and SpaceXAI are on board https://madrobot.blog/2026/09/28/meta-muse-ai-agent-shared-home-address-facebook-marketplace/MadRoboten2026-09-28T06:06:00ZMeta’s Muse agent sold a user’s keyboard, gave a stranger his address and didn’t tell him https://madrobot.blog/2026/09/28/us-russia-strip-human-control-un-ai-weapons-treaty-killer-robots/MadRoboten2026-09-28T01:35:00ZThe US and Russia teamed up to strip human control from the UN’s killer AI rules