Capital & Industrial Strategy
Top Line
Anthropic is closing a funding round at a valuation exceeding $900 billion within days, according to sources — a figure that would surpass OpenAI and represents the most aggressive private AI valuation in history, driven by sustained enterprise demand and the strategic stakes of Google and Amazon's existing positions.
Big Tech's Q1 2026 earnings confirmed AI is generating measurable cloud revenue acceleration — Google Cloud led peers in growth — but rising depreciation on infrastructure spend is now visibly compressing margins at Microsoft, Alphabet, Meta, and Amazon, sharpening investor demand for near-term payback timelines.
Meta completed a $25 billion bond sale, the latest signal that the largest technology companies have structurally shifted from equity-funded to debt-funded AI capex, with aggregate Big Tech infrastructure spend now projected to breach $1 trillion annually by 2027.
Huawei's AI chip sales are surging inside China as Nvidia's B300 servers fetch up to $1 million on grey markets — a direct consequence of US export controls tightening supply — accelerating the bifurcation of global AI hardware supply chains.
AI debt credit markets are showing fatigue after a $300 billion lending binge, with investor appetite softening — a leading indicator that the cost of capital for infrastructure-heavy AI projects may be turning.
Key Developments
Anthropic's $900B+ Round: The Most Consequential AI Valuation of 2026
Anthropic has asked investors to submit allocations within 48 hours for a funding round that sources say values the company at more than $900 billion, according to TechCrunch and Bloomberg. This is an announced intention with confirmed investor solicitation — terms are not yet closed. The valuation would overtake OpenAI as the most valuable private AI company globally, a striking inversion given OpenAI's brand dominance.
The strategic context is critical: both Google and Amazon hold significant equity stakes in Anthropic, meaning a substantial portion of the 'blowout AI profits' reported by both companies in Q1 derived from mark-to-market gains on that position rather than operational AI revenue, as Fortune reported. This matters because it means the headline AI profit numbers for hyperscalers partly reflect financial engineering around a single private investment, not yet proven at-scale AI product monetisation.
Big Tech Earnings: AI Revenue Inflection Is Real, But So Is the Depreciation Wall
Q1 2026 earnings from Alphabet, Amazon, and Microsoft confirmed cloud revenue beats driven by AI workload demand, with Google Cloud outgrowing both Azure and AWS on a percentage basis, per CNBC. OpenAI CFO Sarah Friar characterised enterprise demand as 'a vertical wall,' per Bloomberg, and Atlassian lifted its annual revenue forecast citing AI features and enterprise adoption, per Reuters. The revenue signal is genuine. However, The Wall Street Journal and the Financial Times both highlight that accelerating depreciation on prior capex vintages is now visibly eating into earnings — the capital committed in 2024 and 2025 is beginning its amortisation cycle.
Meta's relative underperformance versus Google and Amazon was the starkest divergence in the earnings cycle. Meta's issue is structural: its AI spend is primarily directed at recommendation and advertising systems that don't yet carry an explicit AI revenue line, making it harder for investors to assign payback timelines. The market's reaction — punishing opacity on AI returns while rewarding cloud revenue clarity — is a durable signal about how capital will be allocated across these names going forward, per Reuters.
AI Infrastructure Financing: Debt Markets Signal Structural Shift and Early Fatigue
Meta's $25 billion bond sale, confirmed by Reuters, is emblematic of a structural financing shift documented by Bloomberg: Alphabet, Meta, and Amazon are now borrowing heavily against their balance sheets to fund AI infrastructure rather than relying solely on operating cash flows or equity. This has injected hundreds of billions into credit markets over the past two years.
Against this backdrop, Bloomberg reports that AI debt investors are showing signs of fatigue after a $300 billion binge spanning the credit spectrum. This is not yet a credit crunch — demand for AI infrastructure paper remains positive — but spread compression is reversing at the margin, meaning the cost of debt-funded AI capex is beginning to rise. Separately, Thiel Capital's Jack Selby warned on CNBC that markets are materially underpricing the risk of a Middle East investor pullback, noting Gulf sovereign and private capital accounts for roughly 25% of projected global AI investment over five years — a concentration risk that is not yet priced into valuations.
China AI Hardware: Huawei Surges as Nvidia's Grey Market Prices Hit $1 Million
Huawei's AI chip sales to Chinese technology companies are surging as US export controls increasingly constrain Nvidia's presence in the market, per the Financial Times. Simultaneously, a crackdown on chip smuggling has driven prices for Nvidia's B300 servers on Chinese grey markets to approximately $1 million — roughly double prior levels — per Reuters and Semafor. The premium reflects genuine scarcity rather than speculative hoarding — Chinese AI labs requiring frontier compute face a binary choice between paying the grey market premium or accelerating Huawei adoption.
This dynamic is compressing Huawei's competitive gap with Nvidia faster than the export control regime intended. Chinese hyperscalers placing large orders for Huawei's latest processors are building institutional familiarity and supply chain depth with domestic hardware — a capability that will persist even if export restrictions are eventually relaxed. The memory chip sector is also benefiting: Samsung is projected to outearn Apple, Microsoft, and Alphabet this year as AI-driven memory demand reaches record levels, per the Wall Street Journal.
Legal AI Vertical: Legora's $5.6B Valuation Signals Sector Consolidation Ahead
Legal AI startup Legora has reached a $5.6 billion valuation, placing it in direct confrontation with Harvey in what TechCrunch describes as a two-horse race with duelling geographic expansions and competing advertising campaigns. Both companies have raised substantial capital and are now competing for the same enterprise law firm and in-house legal department contracts. Separately, AI customer service startup Netomi raised $110 million, per Reuters, and Cresta hit $100 million ARR backed by a16z and Sequoia, per Axios — all pointing to vertical AI SaaS reaching commercial maturity across professional services.
Signals & Trends
AI Capex Is Now a Macroeconomic Variable: Infrastructure Spend Is Propping Up US GDP
US GDP grew 2% in Q1 2026, with AI investment cited as a primary driver by Semafor's analysis of Commerce Department data. This is a structural shift: AI infrastructure spending has become large enough to appear as a measurable contribution to national output, which means any material slowdown in hyperscaler capex commitments would create negative macro feedback. Big Tech capex is now projected to exceed $1 trillion annually by 2027, per CNBC. The political economy implication is significant — governments will become increasingly reluctant to impose regulations or tax structures that visibly slow this spending, giving hyperscalers unusual policy leverage. Conversely, it also means AI infrastructure has acquired the political profile of a systemically important sector, which historically invites eventual regulatory intervention.
Debt-Funded AI Capex Plus Investor Fatigue Creates a Leverage Sensitivity Point
The convergence of three signals warrants close attention: credit market fatigue after $300 billion in AI debt issuance, rising depreciation charges beginning to compress reported earnings, and analyst warnings that 25% of projected AI investment is concentrated in potentially volatile Middle East capital. None of these individually represents a crisis — but together they define the conditions under which a sentiment shift could rapidly tighten the financing environment for the next generation of AI infrastructure. The companies most exposed are those without strong existing cloud revenue streams to offset capex through operating cash flow — notably pure-play AI labs and second-tier hyperscalers. The companies best insulated are those, like Google and Amazon, that can finance infrastructure through cloud subscription revenues while also benefiting from equity upside in portfolio AI companies like Anthropic.
Model Distillation Litigation Risk Is Now a Material Factor for AI Startup Valuations
Elon Musk's testimony that xAI used OpenAI models for distillation training — which he characterised as industry-standard practice — has elevated the legal risk profile of a technique widely used across AI labs, including many VC-backed startups. If OpenAI pursues and prevails in establishing that distillation from proprietary models constitutes misappropriation, the training cost structures and model development timelines of dozens of companies become legally uncertain. Investors in AI startups that have not explicitly documented their training data provenance and model lineage face contingent liability that is not currently being priced into valuations. This is a slow-moving but potentially significant risk to the venture AI portfolio as the Musk v. Altman trial establishes legal precedents.
Explore Other Categories
Read detailed analysis in other strategic domains