Capital & Industrial Strategy

A Guardian investigation found the UK government's multibillion-pound AI investment programme relies on phantom infrastructure and inflated accounting. The flagship Essex supercomputer site announced with fanfare remains a scaffolding yard, whilst investments counted toward national AI capacity turn out to be rented datacentres shipped elsewhere. British startup Nscale raised $2 billion at a $14.6 billion valuation despite minimal operational delivery, appointing former Meta executives Sheryl Sandberg and Nick Clegg to its board as reported by The Guardian and Financial Times.

The investigation reveals successive Conservative and Labour governments counted the same infrastructure multiple times across different announcements, with some investments representing 350,000 percent markups on actual hardware costs. Planning permission for the Essex site was filed only after Guardian inquiries, according to follow-up reporting. Nscale describes itself as building Stargate Norway infrastructure, positioning for government procurement despite the gap between announcements and delivery.

Anthropic filed two federal lawsuits against the Department of Defense challenging its unprecedented designation as a supply-chain risk, claiming the action is unlawful and violates first amendment rights as reported by Wired, Bloomberg, and Financial Times. The company claims the designation could cost billions in revenue as enterprises pause deal negotiations over fear of federal procurement restrictions. A Pentagon official told Bloomberg there is little chance of resuming negotiations following the legal challenge.

Hours after filing, 40 employees from OpenAI and Google including Google DeepMind chief scientist Jeff Dean submitted an amicus brief supporting Anthropic's lawsuit as reported by Wired and TechCrunch. The dispute originated over Anthropic's refusal to remove red lines preventing military use of its Claude chatbot for surveillance targeting Americans, escalating when the Trump administration converted a contract dispute into a federal supply-chain risk designation.

Yann LeCun's AMI Labs raised $1.03 billion in seed funding at a $3.5 billion pre-money valuation to build world models enabling AI to understand physical reality, as reported by Wired, TechCrunch, Bloomberg, and Financial Times. This marks Europe's largest seed round and the veteran researcher's first startup since leaving Meta, where he served as chief AI scientist. Backers include Nvidia, Temasek, and Jeff Bezos according to the Financial Times.

LeCun has long argued human-level AI requires mastering physical world understanding rather than language alone, positioning AMI Labs against the generative pre-training paradigm that powered ChatGPT and similar chatbots. The billion-dollar seed round demonstrates continued investor appetite for fundamental AI research despite market consolidation around existing foundation model providers, suggesting capital remains available for alternative technical approaches if backed by credible scientific leadership.

Microsoft launched a $99 per month AI-focused software bundle integrating Anthropic's Claude models alongside OpenAI technology into its Copilot workplace tools, as reported by Financial Times and Bloomberg. The move marks Microsoft's first major diversification away from exclusive reliance on OpenAI's GPT models for enterprise AI products, positioning Copilot as a multi-model platform rather than a single-vendor channel.

The integration follows Microsoft's $13 billion investment in OpenAI and comes as SoftBank's bet on OpenAI shows signs of strain according to Financial Times reporting on recent share price declines and negative outlook from rating agency S&P. By offering multiple foundation models through a single enterprise subscription, Microsoft reduces customer lock-in concerns whilst maintaining optionality if OpenAI's competitive position shifts.

China's Gen Z retail traders increasingly rely on AI chatbots for investment decisions and are becoming an important driver of capital flows in the world's second-largest economy, according to Bloomberg. Meanwhile, Chinese technology firms including Tencent and Zhipu are rallying on launches of AI agents tapping into OpenClaw, the open-source framework gaining rapid adoption, as reported by Bloomberg.

The convergence of retail traders using AI for portfolio decisions whilst local tech firms build commercial products on open-source AI agents demonstrates how accessible AI tools are reshaping capital allocation beyond institutional investors. Chinese firms moving swiftly to embrace OpenClaw mirrors Western adoption patterns but occurs within different regulatory and competitive dynamics, suggesting open-source AI frameworks enable parallel but distinct market development across geographies.