Capital & Industrial Strategy

The US government designated Anthropic a supply chain risk over undisclosed national security concerns, prompting TechCrunch to question whether the controversy will chill other startups' willingness to pursue defense contracts. Simultaneously, London Mayor Sadiq Khan extended a public invitation for Anthropic to expand in the UK, as reported by the BBC. The divergence illustrates how national security restrictions in one jurisdiction create immediate market opportunities in another, with allied governments competing for AI talent and corporate presence even as they nominally coordinate on technology export controls.

The timing matters strategically: this occurs as defense tech investment has become a mainstream venture category and as European capitals actively court US AI firms frustrated by domestic regulatory or security constraints. The designation's rationale remains opaque, but its market effect is clear — it fragments the transatlantic AI ecosystem and forces startups to choose jurisdictions based on regulatory risk rather than pure economics.

KKR is pursuing a multibillion-dollar sale of data centre cooling company CoolIT Systems, according to the Financial Times, seeking to capitalise on surging valuations for AI infrastructure components. The private equity firm is betting that strategic buyers and infrastructure funds will pay premium multiples for assets that de-risk the physical buildout required to support accelerating compute deployment. Cooling systems represent a critical bottleneck as data centres shift to liquid cooling for high-density GPU clusters, and suppliers with proven deployments at scale command scarcity value.

This exit attempt arrives as Bloomberg reports that AI demand is creating a historic memory chip shortage, with supply constraints now expected to flow through to consumer electronics, automotive, and industrial equipment pricing. The twin pressures of compute infrastructure scarcity and component shortages are reshaping infrastructure investment theses — picks-and-shovels plays are no longer defensive diversification but core AI exposure.

Bloomberg reports that Shenzhen authorities announced new support measures for companies developing tools using OpenClaw, an open-source AI agent framework, driving gains in linked equities. Separately, Bloomberg notes that China's broader 'smart economy' industrial strategy is creating investor conviction around identifiable equity winners in semiconductors, AI, and frontier technology sectors. This represents a coordinated state-directed approach to shaping AI market structure — Beijing is using procurement mandates, subsidies, and regulatory preferences to channel capital toward domestically controlled AI capabilities.

The policy architecture differs fundamentally from Western approaches: rather than attempting horizontal regulation of AI capabilities, China is using vertical industrial strategy to build domestic champions while maintaining open-source positioning where it lacks technological leadership. For global investors, this creates a bifurcated market — Western AI plays centred on proprietary models and cloud services, Chinese plays centred on application-layer deployment and semiconductor self-sufficiency.

Bloomberg reports that AI data centre demand is creating a historic memory chip shortage that will make phones, cars, and consumer electronics more expensive. The constraint is not cyclical oversupply correction but structural undersupply driven by exponential AI compute requirements. Meeting this demand will require unprecedented capital expenditure in fabrication capacity, and even then supply may not keep pace with model training and inference scaling trajectories. The shortage is already affecting spot pricing for HBM (high-bandwidth memory) and DDR5, with allocation increasingly directed toward hyperscale AI buyers willing to pay premium pricing and commit to multi-year offtake agreements.

The cost implications extend beyond data centres: automotive manufacturers reliant on advanced driver-assistance systems, consumer electronics companies embedding on-device AI, and industrial automation vendors all compete in the same supply chain. This creates a margin squeeze for hardware companies unable to pass through component cost increases, potentially reshaping competitive dynamics in sectors previously insulated from AI economics.