Hyperscaler cloud revenues may top $1 trillion annually by 2030 on AI demand

Compute capacity is becoming an economic resource as AI and blockchain-based digital assets increasingly converge.Hyperscaler cloud revenues could exceed $1 trillion annually by 2030 as demand grows for the computing resources needed to train and run artificial intelligence (AI) systems, according to an ANI report.Compute capacity is becoming an economic resource as AI and blockchain-based digital assets increasingly converge.‘Compute, the processing capacity required to train and run AI systems, is becoming an increasingly important economic resource,’ the report added.Citing analyst estimates, the report said hyperscaler cloud revenues could surpass $1 trillion a year by 2030.

It added that standardised claims on computing capacity could emerge as a digital asset use case for financing and programmable settlement as AI agents become more capable and persistent.Liquid compute markets could allow autonomous software agents to source, optimise and pay for hardware infrastructure directly.AI meets digital assetsThe report highlighted a structural connection between AI and digital assets, describing AI as ‘machine-native intelligence’ and digital assets as ‘machine-native money’.The relationship becomes particularly important with the emergence of agentic AI—systems capable of planning and executing multistep tasks towards a defined objective by interacting with external tools and infrastructure with limited human intervention.Blockchains can provide programmable infrastructure connecting such intelligence with economic activity, the report said.The report also pointed to similarities in the tokenisation models used by AI and digital assets.

Large language models (LLMs) break language into tokens for numerical evaluation, while distributed ledgers register economic entitlements as tokens that can be transferred in a machine-verifiable manner.Agentic AI could drive programmable paymentsAccording to the report, agentic commerce requires programmable payment infrastructure.

At the same time, conventional mechanisms such as automated clearing houses and card networks have onboarding requirements and cost structures that make them less suited to continuous, low-value microtransactions.Autonomous software systems could instead use stablecoins and specialised transfer protocols such as x402 and ACP, according to the report.‘The structured, machine-readable representations created through LLM and blockchain tokenization can give AI agents a more direct interface with programmable assets, while stablecoins and protocols such as x402 may support high-frequency, low-value, always-on transactions,’ it said.The report, however, said the ecosystem remains in its early stages, with liquidity across compute claims and transaction volumes generated by AI agents still modest.As machine autonomy increases, digital assets, tokenised real-world assets and base settlement cryptocurrencies could become components of machine-to-machine financial activity.