Macquarie Group Ltd. and Singapore's sovereign wealth fund GIC Private Ltd. have entered into a strategic partnership with frontier artificial intelligence lab Anthropic PBC to develop, own and lease data center infrastructure, with Anthropic serving as the anchor tenant. The collaboration, operating under the name Theseus Infrastructure, will be owned and funded by funds managed by Macquarie Asset Management alongside GIC, with the facilities then leased to Anthropic under long-term arrangements.
The partnership underscores the escalating capital demands of the digital infrastructure boom. While the parties did not disclose a total investment figure, the joint announcement noted that "the planned developments will require significant capital investment." According to consulting firm estimates, global spending on digital infrastructure development could reach $7 trillion by 2030, highlighting the scale of investment needed to support the rapid expansion of AI-driven computing capacity.
For GIC, the deal adds to a growing portfolio of data center investments, which includes stakes in Vantage Data Centers and a joint venture with the Canada Pension Plan Investment Board and Equinix. Macquarie, meanwhile, has committed tens of billions of dollars to data center projects over the years, including its acquisition of Aligned Data Centers in 2018—sold this year for $40 billion—as well as investments in Applied Digital and AirTrunk.
Anthropic said in a statement that demand for its AI assistant Claude "continues to grow rapidly across businesses, developers, and consumers, and meeting that demand requires significant new computing power." The company added: "Anthropic is partnering with Macquarie Asset Management and GIC, drawing on Macquarie's global expertise developing, financing, and operating large-scale digital infrastructure, and GIC's deep experience investing in infrastructure around the world."
The move reflects a broader trend of AI developers securing dedicated infrastructure capacity through strategic partnerships with financial investors, as the race to scale advanced models intensifies and traditional data center supply remains constrained.