Capital & Industrial Strategy

Anthropic is in early-stage talks to raise at least $30 billion in fresh financing at a valuation above $900 billion, according to Bloomberg. This would be its largest funding round yet and would place it within striking distance of public tech giants by market cap — a remarkable position for a company that was widely regarded as a distant second to OpenAI just eighteen months ago. These are confirmed talks at an early stage, not a closed deal; terms and investors have not been finalised.

The fundraising narrative is underpinned by hard commercial data. Ramp's expense analytics across its client base show 34.4% of businesses now pay for Anthropic services versus 32.3% for OpenAI — a reversal that TechCrunch describes as the first time Anthropic has led on this metric. Simultaneously, Anthropic launched a small business offering, explicitly targeting the 36 million US small businesses, indicating a deliberate downmarket expansion beyond its enterprise-and-developer core. The strategic logic is straightforward: having won on coding and enterprise reliability, Anthropic is now using that credibility to broaden its customer base before OpenAI can lock in SMB relationships through its own product push and the TPG-backed consulting joint venture.

Cerebras Systems raised $5.55 billion at $185 per share — above its already-elevated expected range — valuing the company at $40 billion according to the Financial Times. This is a confirmed, priced deal. The offering is notable not just for its size but for its timing: Cerebras had previously faced a delayed IPO path partly due to regulatory scrutiny over its relationship with a Middle Eastern backer. The clean execution above range signals that those concerns have been resolved to investor satisfaction and that demand for non-Nvidia AI silicon stories is deep enough to support a $40 billion public valuation.

The Cerebras listing is being read by Wall Street as the opening act of a larger AI IPO wave expected later in 2026, per CNBC. Concurrently, hedge funds are reporting their best months in decades from AI hardware stock-picking, with firms like Point72 heavily concentrated in chips and related equipment according to the Wall Street Journal. SK Hynix is approaching a $1 trillion market cap, and Tower Semiconductor signed $1.3 billion in AI chip deals. The risk flagged by Reuters is that the semiconductor trade is becoming crowded enough to threaten the broader equity rally if sentiment reverses.

Blackstone executed two significant AI infrastructure capital raises in parallel. Its Digital Infrastructure Trust raised $1.75 billion in a US IPO to acquire data centers, per Bloomberg — a confirmed closed deal. Simultaneously, its portfolio company AirTrunk is marketing a $2.3 billion loan to fund a Malaysia data center project, per Bloomberg. The Malaysia financing is a confirmed marketing process, not yet a closed deal. Together, these moves represent Blackstone deploying alternative asset management infrastructure — REIT structures, leveraged loans — to channel institutional capital into AI infrastructure at a scale and speed that pure tech corporate balance sheets cannot match.

The Southeast Asia dimension of the AirTrunk deal is strategically significant. Malaysia has emerged as a preferred jurisdiction for US-aligned AI infrastructure investment given its political stability, available land, and cooling capacity, and as a partial workaround to restrictions on building advanced AI compute in geopolitically sensitive locations. This is capital following government industrial strategy signals, not leading them.

Reuters reports exclusively that Microsoft is actively eyeing startup acquisition deals as it plans for reduced dependence on OpenAI, confirming a strategic shift that has been speculated for months. Simultaneously, the Musk v. Altman trial produced testimony from senior Microsoft executives acknowledging that the company feared being overly dependent on OpenAI — a notable public admission from a company that has staked much of its Copilot and Azure AI strategy on that partnership, per CNBC. The Reuters report describes an active deal-scouting posture; no specific acquisitions have been announced or confirmed.

This diversification imperative is occurring against a backdrop where OpenAI itself is expanding its commercial surface area — the TPG-backed consulting joint venture, the small business push — in ways that could eventually put it in direct competition with Microsoft's own enterprise services. TPG's David Trujillo described the OpenAI consulting venture as a traditional services play for the firm, per Bloomberg, framing it as accelerating enterprise adoption rather than a departure for OpenAI. But the strategic tension between Microsoft and OpenAI is now openly acknowledged at the executive and legal level.

Alibaba and Tencent both reported first-quarter results that missed revenue expectations, but investor reactions were muted to positive as management framing on AI investment dominated. Alibaba stated AI spending will exceed original goals and explicitly said margin is 'secondary,' per Reuters. Tencent's gaming and cloud segments provided cushion as AI investment weighed on margins. Bloomberg notes that investors are explicitly looking past near-term financials to long-term AI platform positioning — a valuation stance that mirrors early hyperscaler adoption cycles in the US.

The geopolitical dimension sharpened significantly with Nvidia CEO Jensen Huang's last-minute addition to President Trump's Beijing trip, raising market expectations of a deal to resume H200 chip sales to China, per Reuters and Semafor. This remains speculative — no deal has been announced — but Huang's presence as a 'bargaining chip' in diplomatic negotiations underscores how AI semiconductor access has become an explicit instrument of US-China trade policy.