The morning briefing.

Capital and compute. An Epoch AI analysis suggests major hyperscalers like Microsoft and Amazon may soon struggle to fund their rapidly expanding AI infrastructure solely from operating cash flow. Their AI spending is growing at 70% annually, far outstripping the 23% rise in cash flow, potentially requiring external funding by Q3 2026. This financial pressure is also driving hardware component suppliers, as seen with Kingboard Holdings raising $1.5 billion to boost PCB capacity for AI servers. The demand for AI hardware continues to outstrip current production capabilities.

Frontier models. The competitive landscape for AI models is intensifying, with GLM-5.2 emerging as a leading open-weights model, particularly noted for long-horizon tasks and frontend coding. This comes as Mistral's CEO emphasizes open-source AI's independence from state control, following recent regulatory scrutiny on frontier models. Meanwhile, ChatGPT's market share has dipped below 50% for the first time, indicating growing adoption of alternatives like Gemini and Chinese AI models such as DeepSeek among American users seeking cost-effective solutions.

AI in daily life. AI is increasingly integrated into consumer applications and robotics, transforming how users interact with technology. Pinterest has launched 'Ask Pinterest,' an experimental AI shopping app that provides conversational recommendations and inspiration. In transportation, Uber plans to deploy premium robotaxi services in Houston by 2027, utilizing Lucid EVs equipped with Nuro's self-driving system. Robotics innovation is also challenging traditional forms, with Genesis AI introducing Eno, a "general-purpose" humanoid robot designed for capability rather than human appearance.

Regulatory oversight. Global leaders are actively engaging in discussions around AI governance and safety, with Trump and other world leaders joining major AI companies like OpenAI, Anthropic, and Google at the G7 summit to address frontier AI risks and infrastructure. Concurrently, China is accelerating its own AI safety cooperation efforts, publishing a global governance whitepaper. On the data privacy front, Apple is reportedly collecting extensive analytics data, including every tap, to power its new App Store personalized recommendations, raising concerns among security researchers.

European tech funding. European startups are securing significant funding rounds to advance AI and cybersecurity solutions. Barcelona-based NeuralTrust raised €17.2 million in seed funding, reportedly the largest cybersecurity seed financing for an EU company, to secure and govern enterprise AI agents. Similarly, Flagright, an AI operating system for financial crime compliance, closed a $12.5 million Series A round. These investments highlight Europe's growing focus on developing robust AI infrastructure and security measures.