[DigitalToday reporter Jinju Hong] BlackRock Chief Executive Larry Fink (래리 핑크) said the spread of artificial intelligence could make computing resources themselves a new asset class.

BeInCrypto, a blockchain media outlet, reported on May 6 that Fink said at a Milken Institute event a futures market could emerge for buying and selling raw computing power.

Fink defined compute not as a simple cloud service but as a scarce resource that financial markets can price. He compared compute to commodities such as energy or agricultural products, and explained that companies could manage AI infrastructure costs with structured futures contracts, much as they hedge fuel prices. Brookfield CEO Bruce Flatt (브루스 플랫) also appeared on stage at the event.

Fink said, "The new asset class is buying compute futures." The remark is seen as a plan to bundle physical infrastructure such as data centres, semiconductors, memory and power into underlying assets for financial products as AI investment expands. For institutional investors, it could be a way to lock in the costs of chip and power capacity needed to run models.

He also stressed that current supply conditions are not keeping up with AI demand. The United States does not yet have enough chips, memory and power facilities to handle the expected AI workload. Fink said, "We don't have enough power, we don't have enough compute, we don't have enough chips."

Fink also pushed back against claims that AI investment is overheating. He said he does not accept views in parts of the market that it is a bubble, and that demand continues to outstrip supply. He also said that although cloud companies and large semiconductor firms are continuing to raise large amounts of funds, that alone may not be enough to meet global demand for expanding data centres. He warned that the sector could face a situation in which capital itself becomes scarce.

BlackRock's moves also align with that view. Fink said BlackRock plans to announce within this week a partnership with an unnamed hyperscale cloud operator. The deal is expected to be an opportunity to deploy more of BlackRock's $13.9 trillion in assets into AI infrastructure. The focus is also shifting toward directly holding stakes in physical infrastructure, beyond simply providing finance. Fink did not disclose the counterparty's name before the official announcement.

The market is focusing on the need to resolve standardisation issues before a compute futures market can be created. For exchanges to list such products, they must decide what to use as the standard unit, reflecting differences in hardware by generation and constantly changing AI workloads. The industry has yet to reach a consensus on how to measure computing performance and convert it into contract units.

Even so, Fink's remarks show the possibility that capital flows around AI infrastructure could shift into a more concrete market for physical assets. With demand rising simultaneously for power, semiconductors and data centres, BlackRock views compute as a physical investment target that long-term capital can hold. The key questions are whether the industry can set a standard unit for compute and whether a derivatives market will open based on it.