Bitcoin Miner Cango Sells Millions in Bitcoin to Cut Debt and Fund AI Expansion

Bitcoin Magazine Bitcoin Miner Cango Sells Millions in Bitcoin to Cut Debt and Fund AI Expansion Cango (CANG) said it sold 4,451 Bitcoin over the weekend for net proceeds of about $305 million as the company moves to strengthen its balance sheet and support a shift into artificial intelligence infrastructure. The Dallas-based Bitcoin miner announced Monday that the transaction was settled directly in Tether’s USDT stablecoin. The company said the full amount of the proceeds was used to partially repay a Bitcoin-collateralized loan. Cango said the sale followed a review of market conditions and was approved by its board of directors. The company framed the move as a balance-sheet adjustment aimed at reducing leverage rather than a retreat from its mining business. The company’s stock is currently down 9%. “The divestment of a portion of the Company’s Bitcoin holdings was executed to strengthen its balance sheet and reduce financial leverage,” Cango said in its statement. The company said the debt reduction provides greater capacity to fund its strategic expansion into AI compute infrastructure. Cango is pursuing a plan to build an integrated energy and AI compute platform by using its grid-connected mining sites to provide distributed computing services for the AI industry. The bitcoin miner said its approach will roll out in phases. The first stage will deploy modular, containerized GPU compute nodes across existing sites. The company said it plans to offer inference capacity for small and medium enterprises, a segment it described as underserved. A later phase will focus on building a software orchestration platform to unify distributed compute resources across its global footprint. Cango’s AI and bitcoin miner pivot As part of the AI push, the bitcoin miner announced the appointment of Jack Jin as chief technology officer of its AI business line. The company said Jin previously worked at Zoom Communications, where he led deployments of multi-node GPU clusters s...