The evening briefing.

AI Policy & Regulation. The US government ordered Anthropic to block foreign access to its Fable 5 and Mythos 5 models, citing risks of diversion to foreign military intelligence. This move, which some analysts suggest was not about an AI jailbreak but rather a broader policy decision, highlights growing national security concerns around advanced AI capabilities. The incident underscores the increasing scrutiny and direct intervention by governments in the development and deployment of frontier AI.

Frontier Model Challenges. Anthropic is facing a new lawsuit alleging misleading users about soaring AI costs and usage limits for its Claude Max model. This comes as smaller AI labs, like Cohere with its North Mini Code, aim to offer developers more control and transparency, appealing to those who find frontier models from Anthropic and OpenAI less accessible. Meanwhile, Google DeepMind's Gemma 4 models are now available on Amazon Bedrock, expanding access to open-weight models focused on intelligence-per-parameter.

AI Infrastructure and Investment. Nvidia plans to raise at least $20 billion in its first debt sale since the start of the AI boom, signaling continued aggressive investment in the sector. However, a Gartner report warns that data center growth could halt by 2030 due to power demands, potentially creating an AI bottleneck not related to silicon but to energy access. Separately, IMEC's new 6G chip breakthrough addresses an efficiency issue, aligning with Nvidia CEO Jensen Huang's vision for AI in telecom.

Digital Platform Governance. The UK is moving to ban social media for children under 16 and may impose overnight curfews, a bold policy aiming to protect minors online. Concurrently, Meta is rolling out a new "AI Mode" for Facebook search that pulls information from public posts across its platforms, raising questions about data usage and content moderation. This highlights the ongoing tension between platform innovation and regulatory efforts to control digital spaces.