Capital & Industrial Strategy

Nvidia disclosed in filings that it will spend $26 billion to build open-weight AI models, according to Wired, a move that positions the chip giant to compete directly with model developers like OpenAI, Anthropic, and DeepSeek rather than simply supplying them with infrastructure. Simultaneously, Nvidia struck a $2 billion deal with AI cloud provider Nebius, according to the Financial Times, continuing a dealmaking pattern where the company deploys its enormous cash reserves to finance its own customers. This dual strategy — building models and funding infrastructure buyers — represents a fundamental shift from pure hardware supplier to vertically integrated AI platform provider.

The timing is significant: as model development costs plateau and open-weight alternatives gain traction, Nvidia appears to be hedging against a future where differentiation comes from model performance rather than chip performance. By funding cloud providers who buy its chips while also building models that require those chips, Nvidia is constructing a self-reinforcing ecosystem. The $26 billion commitment to model development exceeds the annual revenue of most AI startups, suggesting Nvidia views this as a core business line rather than an experimental side project.

AI coding startup Cursor is in talks for a funding round valuing the company at approximately $50 billion, according to Bloomberg, nearly double the valuation it secured last fall. This follows Replit's announcement of a $400 million raise at a $9 billion valuation according to TechCrunch, tripling its $3 billion valuation from just six months ago, with the company targeting $1 billion in annual recurring revenue by year-end. Swedish competitor Lovable, with only 146 employees, claims it added $100 million in revenue in February alone and crossed $400 million in ARR, according to TechCrunch.

These valuations reflect a market consensus that AI will fundamentally reshape software development, with the winning platforms capturing massive value as they replace traditional development workflows. The revenue acceleration is extraordinary even by AI standards — Lovable's claimed February performance would represent $1.2 billion annualised from a single month. However, the sustainability of this growth remains unproven: revenue multiples at these levels assume not just continued adoption but successful monetisation of usage that may currently be experimental or subsidised.

Netflix will pay up to $600 million for InterPositive, the AI filmmaking company founded by Ben Affleck, according to Bloomberg and TechCrunch, making this one of the streaming giant's largest acquisitions ever. The deal signals Netflix's strategic pivot toward owning AI-driven production infrastructure rather than simply licensing content or tools. InterPositive has positioned itself as building AI systems specifically for feature film and episodic production workflows, rather than general-purpose generative video tools.

The acquisition price — potentially reaching $600 million depending on performance milestones — suggests Netflix sees AI production tools as a strategic capability worth building or buying at almost any cost. This likely reflects internal analyses showing that AI-assisted production could dramatically reduce the $17 billion Netflix spends annually on content while maintaining or improving output quality. The deal structure, with its performance-based upside, indicates Netflix is betting on InterPositive's technology delivering measurable production cost savings within a defined timeframe.

Salesforce's $25 billion bond offering drew lukewarm demand and required significant yield concessions, according to Bloomberg and the Financial Times, with investors citing concerns over the company's debt-funded share buyback programme and broader worries about software firms' vulnerability to AI disruption. The weak reception marks a notable shift from the easy money software companies enjoyed in previous years. Separately, Atlassian announced 1,600 layoffs — 10% of its workforce — according to Bloomberg and the Financial Times, explicitly citing the need to restructure around AI, while Oracle set aside an additional $500 million for restructuring costs as it touts productivity gains from AI coding tools, according to the Financial Times.

The bond market's scepticism reflects growing concern that legacy software companies face existential risk from AI-native competitors who can undercut on price due to dramatically lower development costs. Investors are essentially pricing in a future where today's software giants face compressed margins as AI agents automate away the work their products currently charge for. The restructuring charges at Oracle and Atlassian suggest even the companies themselves recognise their current cost structures are unsustainable in an AI-augmented world — though the productivity gains cited may be more aspirational than realised, given reports from companies like Amazon where AI tools are creating more work rather than less.

The UK announced that water utility takeovers will face mandatory national security screening while commercially available AI systems will be removed from the investment screening list as part of a red-tape reduction drive, according to the Financial Times. This policy divergence — tightening scrutiny on critical infrastructure while loosening it on AI — reflects a gamble that AI investment needs regulatory certainty more than oversight. Meanwhile, Microsoft announced an Africa AI adoption push according to Bloomberg, explicitly framing it as competition with China's DeepSeek, while Bridge Data Centres announced plans for up to $3.9 billion in AI investment in Singapore, according to Bloomberg.

The Guardian reported that 11 African governments have collectively spent over $2 billion on Chinese-built AI-powered surveillance systems according to The Guardian, creating a geopolitical competition where Western companies are racing to offer AI infrastructure before Chinese systems become entrenched. The UK's decision to exempt AI from mandatory screening while tightening rules elsewhere suggests governments are treating AI investment as a strategic priority that justifies regulatory forbearance — a stance that may prove shortsighted if those systems later become national security concerns.