Meta Platforms is in early-stage talks to lease computing power to AI company Anthropic in a deal that could be worth up to $10 billion over two years, according to a source familiar with the matter.
The discussions represent a significant strategic shift for Meta, which has historically focused on advertising as its primary revenue source. Under the proposed arrangement, Anthropic — the creator of the Claude AI model family — would pay Meta in monthly installments over the two-year period. The terms remain subject to change, and either company would be able to exit the agreement early.
Anthropic initially proposed the deal in June as it prepares for a highly anticipated initial public offering. The company had confidentially filed for its IPO earlier that month. Meta is now considering the proposal, though talks have become complicated since Meta does not yet have an established business selling its computing capacity.
At Meta's shareholder meeting in May, CEO Mark Zuckerberg said entering cloud computing was "definitely on the table," noting that firms were approaching Meta "almost every week" to buy access to its AI models or spare computing power. Earlier in July, Bloomberg News reported that Meta was building a cloud business to sell excess computing capacity and host AI models for developers.
The potential agreement mirrors a strategy recently pursued by Elon Musk's SpaceX, with whom Anthropic struck a deal in May to tap the full computing power of its Colossus 1 data center in Memphis, Tennessee.
Shares of Meta pared losses after the news and closed down more than 2% amid a wider tech selloff on Friday, recovering slightly in extended trading. The deal would help Meta compete with neocloud firms such as CoreWeave and Nebius, as growing adoption of advanced AI tools boosts the need for computing capacity across the industry.
Discussions remain in their early stages and may not result in a finalized agreement, according to the source.
