Alphabet raises $80bn in equity to fund AI infrastructure expansion
The US tech giant cites unprecedented demand for its AI solutions as it scales capital expenditure to $180-190bn this year, aligning with a broader industry shift where underinvestment is viewed as an existential risk.
Alphabet, the parent company of Google, has announced plans to sell $80bn worth of shares to finance the rollout of artificial intelligence infrastructure. The fundraising initiative includes a specific agreement to sell $10bn of stock to Berkshire Hathaway, the conglomerate led by Warren Buffett. The remaining $70bn will be raised through $30bn in underwritten offerings and $40bn in staggered sales on the open market.
The company stated that the equity offerings are necessary to meet what it described as unprecedented customer demand for its AI solutions. In a statement, Alphabet noted that demand from both enterprises and consumers is currently exceeding its available supply. The firm indicated that scaling these investments is essential to expand its foundational infrastructure and support significant growth opportunities ahead.
Alphabet’s shares fell approximately 1 per cent in after-hours trading following the announcement. The company, which has a market capitalisation of more than $4.5trn, reported that its capital expenditure is projected to reach $180-190bn for the current year. Expectations for significant increases in spending are set for 2027.
The move aligns with a broader trend among US technology giants, including Microsoft, Amazon, and Meta, who are collectively expected to spend approximately $800bn on AI-related capital investment in 2026, according to an analysis by Goldman Sachs. Troy Hooper, co-head of equity capital markets for the Americas at Mergermarket, described the funding plans as evidence of the intense competition to lead the AI buildout.
Hooper noted that for hyperscalers, compute capacity is a direct driver of future revenue. He explained that by leveraging equity, Alphabet is securing permanent capital rather than burdening a balance sheet already absorbing record capital expenditure. He added that US tech giants view underinvestment in AI as an existential risk, while overinvestment is merely expensive.
The logic, according to Hooper, is that ownership at scale lowers the marginal cost of training advanced models and builds a competitive moat that smaller rivals will struggle to match. He concluded that the winners of the AI era will be determined not just by algorithms, but by which companies own the largest and most efficient compute platforms.
