Nvidia has suspended some agreements under its AI cloud credit-support program less than two months after launching the initiative in July, according to people familiar with the matter cited by The Wall Street Journal. The program was designed to back AI cloud service providers by guaranteeing revenue if they could not secure enough customers for their GPU capacity. The arrangement, known as the AI Computing Power Partnership Program, involved typical six-year agreements totaling $36 billion in commitments, Nvidia disclosed this week. Under the proposed structure, Nvidia would rent unused GPU computing power itself and, once provider revenue rose above a base threshold covering chip depreciation, data center, and personnel costs, Nvidia could take 50% of the excess revenue. The report said some Nvidia employees were concerned the model could draw antitrust scrutiny, particularly over how much influence the company could exert on customers' business operations. Nvidia may revise the program in the future or fold it into other projects.