Anthropic’s $30 Billion Question: How Much Capital Can AI Really Absorb?
Anthropic’s latest fundraising ambitions mark a new stress test for AI infrastructure investment. According to reports, the Claude maker is in talks to raise at least USD 30 billion (approx. RM138 billion) at a valuation above USD 900 billion (approx. RM4.14 trillion), excluding the new capital. That would more than double the roughly USD 380 billion (approx. RM1.75 trillion) valuation reportedly attached to a prior USD 30 billion (approx. RM138 billion) round in February. Unlike traditional software funding, this Anthropic funding round looks more like a capital-intensive infrastructure raise, aimed at securing fuel, power, chips, and long-term cloud capacity rather than just hiring and marketing. At this scale, investor interest alone does not guarantee cash in the bank; no term sheet has been signed, and the company has declined to comment. Yet the signal is clear: frontier AI valuation is increasingly defined by access to compute rather than purely by user growth or revenue momentum.
Cloud Compute Demand and the $2 Trillion AI Backlog
Anthropic’s capital needs make sense only when viewed against unprecedented cloud compute demand. Contracts involving Anthropic and OpenAI are reportedly now worth over half of the USD 2 trillion (approx. RM9.2 trillion) in revenue backlogs at major cloud providers. In Google’s case, Anthropic alone accounts for more than 40% of the cloud division’s disclosed revenue backlog, underpinned by a commitment to spend USD 200 billion (approx. RM920 billion) over five years on Google Cloud and chips. At the same time, Anthropic has tied itself closely to AWS through a strategic deal that includes more than USD 100 billion (approx. RM460 billion) in planned spending over the next decade and access to as much as five gigawatts of Trainium-based capacity. These long-dated contracts show how AI infrastructure investment is locking in future demand and pushing cloud backlogs to historic levels.
Supercomputing Arms Race: From Shared Cloud to Dedicated AI Infrastructure
Anthropic’s aggressive spend commitments highlight a broader shift: frontier AI companies are racing to convert capital into dedicated supercomputing infrastructure. Industry analysts note that access to advanced supercomputers, AI chips, and scalable cloud environments has become one of the most important strategic assets in the digital economy. Modern large language models demand extraordinary processing power and energy during training, pushing developers away from traditional shared cloud setups toward bespoke, long-term infrastructure partnerships. Anthropic’s expanded access to high-performance computing, including gigawatt-scale capacity from major cloud providers, is emblematic of this trend. The company’s AI infrastructure investment strategy is less about short-term efficiency and more about securing scarce compute before rivals do. As supercomputing build-outs accelerate globally, the companies that can lock in capacity at scale may enjoy durable advantages in model performance, reliability, and the ability to support enterprise AI agents and automation platforms.
Coke vs. Pepsi: Anthropic and OpenAI in the Trillion-Dollar Cloud Compute Race
The contest between Anthropic and OpenAI increasingly resembles a Coke vs. Pepsi battle for AI-era dominance, with both companies preparing for potential mega-IPOs and racing to demonstrate massive compute ramps to investors. Their contracts account for more than half of the USD 2 trillion (approx. RM9.2 trillion) cloud backlog at hyperscalers such as AWS, Azure, and Google Cloud, underscoring how much the AI boom still hinges on two cash-burning startups. Anthropic is aligning deeply with AWS and Google’s TPU and server capacity, while OpenAI’s modified agreement with Microsoft reaffirms its own tight integration with Azure. The frontier AI valuation landscape now depends heavily on who can secure and efficiently deploy these multi-gigawatt data center commitments. In this environment, cloud providers and AI labs are mutually dependent: labs need capacity to meet AI agent demand, and clouds need lab contracts to justify their capital-intensive expansions.
Employee Windfalls and the New AI Wealth Model
Behind the headline valuations and cloud commitments lies a parallel story of employee wealth creation reminiscent of earlier tech booms. OpenAI’s October 2025 internal share sale allowed more than 600 current and former workers to sell USD 6.6 billion (approx. RM30.4 billion) worth of equity without going public, reportedly turning about 75 employees into USD 30 million (approx. RM138 million) sellers each. Subsequent testimony suggested co-founder Greg Brockman’s stake might be worth roughly USD 30 billion (approx. RM138 billion). These figures show how frontier AI companies can generate vast private-market wealth even before IPOs. As Anthropic pursues its own supersized funding, similar outcomes are likely for its staff and early backers. However, such concentrated gains also raise questions about sustainability, liquidity, and how much private capital the AI boom can realistically absorb before public markets, regulators, and customers demand more transparency and tangible returns.
