🏛️ Trump calls Jensen Huang, dismisses AI guardrails as hoax 🏛️
Trump's live call with Nvidia's CEO signals a firm deregulatory stance on AI access.

President Trump called Nvidia CEO Jensen Huang live on stage at the All-In Summit, publicly dismissing fears about artificial intelligence as "a hoax." He argued that guardrails on AI development are unnecessary and that slowing down would hand China a decisive competitive edge—this shapes near-term policy expectations for the entire industry. The remarks reinforce that neither the US nor China intends to agree to any mutual AI slowdown, with direct implications for both domestic legislation and international governance negotiations. Nvidia's continued centrality to the geopolitical AI landscape was on full display, with Huang appearing alongside the President in one of the most high-profile AI policy moments of the year.

📜 Microsoft ships 38-page Humanist AI Code of Conduct 📜
Microsoft's new regulation framework could set the standard for AI model conduct from 2027.

Microsoft released a 38-page draft "Humanist AI Code of Conduct" for its MAI models, built around CEO Mustafa Suleyman's vision of keeping AI subordinate to humans. The core principle: people matter more than AI. If completing a task requires breaking the code, the model must simply refuse. Models must accept being shut down or corrected, keep reasoning auditable, never tamper with their own logs, and never claim consciousness or feelings. Microsoft explicitly rejects "model welfare" and AI legal personhood. Hard lines are drawn around CBRNE weapons, cyberattacks, and non-consensual deepfakes—the Code outranks both users and enterprise deployers, a strong signal favoring self-regulation over government oversight. The document is open to public comment for six weeks, with the revised version set to guide training from 2027.

🧪 Google Labs rolls out Dreambeans AI experiment 🧪
Google's latest Labs experiment explores cross-app AI integration for consumer use.

Google Labs has released Dreambeans, a new experimental tool that surfaces suggestions by pulling context from multiple apps. While details remain limited, the rollout signals Google's continued investment in consumer-facing AI products through its Labs incubator—historically a testing ground for features that later ship inside core products like Search, Workspace, and Android. Dreambeans appears to focus on cross-app intelligence, potentially letting an AI layer integrate and surface relevant information across different applications. The experiment has already attracted attention across AI tool trackers and developer communities, adding to the steady cadence of lightweight AI experiments from Google aimed at exploring novel interaction patterns that could influence future product distribution.

INTERESTING TO KNOW

📸 OpenAI buys Glass Imaging for over $300M 📸

OpenAI has quietly acquired Glass Imaging, a startup developing AI-powered smartphone cameras capable of producing DSLR-quality images, for over $300 million. The acquisition feeds directly into OpenAI's hardware integration effort with former Apple design chief Jony Ive, signaling that camera capabilities will be central to its upcoming device. Founded in 2019 by two former Apple engineers, Glass Imaging represents one of OpenAI's largest known acquisitions—underscoring its ambition to control the full stack from model to device.

🔄 Apple lets Siri swap to Claude or ChatGPT 🔄

Hidden iOS 27 and macOS frameworks reveal that Apple has built a "Model Delegation" system enabling full Siri replacement by Claude or GPT-5.6. An external model can handle natural-language requests and hand actions like Reminders or Messages back to Siri, all while preserving its UI and voice. The rollout was developed partly under EU Digital Markets Act pressure to open default assistants to competitors—a regulation-driven shift that could reshape competitive dynamics across over a billion Apple devices.