Tom Lee has intensified his case for Ethereum as a core settlement network for finance and artificial intelligence, arguing that autonomous AI agents will need a trusted, programmable layer to authenticate instructions, execute transactions and settle value without relying on closed platforms.Lee, chairman of BitMine Immersion Technologies and co-founder of Fundstrat Global Advisors, wrote on X that “ETH remains the future settlement layer of finance and AI”, framing weakness in crypto market sentiment as part of a broader cycle rather than a challenge to Ethereum’s long-term utility. His remarks add weight to a growing debate over whether blockchain networks can provide the economic infrastructure needed for machine-to-machine commerce.
BitMine has become one of the most aggressive corporate buyers of Ether after shifting from its earlier Bitcoin-mining focus to an Ethereum treasury strategy. The company said this week that its Ether holdings had reached 5.28 million tokens, valued at about $11.5 billion at the time of disclosure. That position represented roughly 4.37% of Ethereum’s 120.7 million token supply, putting BitMine close to its stated target of owning 5% of all ETH.
The company has also expanded its staking operations, with about 4.7 million ETH deployed to generate yield. Annualised staking revenue was estimated at $289 million, based on a seven-day annualised yield of 2.80%. The figures underscore how Ethereum treasury companies are moving beyond passive balance-sheet exposure and turning holdings into recurring revenue streams through network participation.
Lee’s argument rests on the view that the next phase of AI will not be confined to chatbots or enterprise copilots. Agentic AI systems are designed to act independently, request services, purchase data, pay for computing resources, execute contracts and interact with other software agents. That model requires payment rails capable of handling high-frequency, low-value transactions with verifiable settlement and programmable conditions.
Ethereum’s supporters say the network is well placed for that role because of its smart-contract architecture, large developer base, stablecoin liquidity and growing layer-2 ecosystem. Stablecoins already account for a large share of blockchain transaction activity, and Ethereum remains a major venue for tokenised dollars, decentralised finance and collateral markets. The same infrastructure, Lee argues, could serve AI agents that need to pay for cloud inference, API calls, data access or digital services.
The case is not uncontested. Ethereum still faces questions over transaction costs, throughput, user experience and competition from faster chains and specialist layer-2 networks. Autonomous AI systems may also require identity, reputation, compliance and dispute-resolution layers that are not fully mature. Academic work on agent economies has highlighted the potential of blockchain-based settlement, but it has also noted that existing systems were built mainly for human-led financial transactions, not large volumes of machine-generated interactions.
Regulation remains another constraint. Stablecoin legislation, tokenisation rules and securities oversight in the United States and other major markets will shape whether institutions can build large-scale settlement systems on public blockchains. Lee has pointed to proposed US market-structure reforms as a potential catalyst, while institutional investors continue to monitor whether clearer rules could unlock more Ethereum-linked financial products.
BitMine’s own strategy has attracted both enthusiasm and scepticism. Supporters compare its Ethereum accumulation to Strategy’s Bitcoin treasury model, arguing that concentrated corporate holdings can give equity investors leveraged exposure to digital assets. Critics warn that the approach can magnify volatility, particularly when companies issue shares or take market risk to accumulate crypto. BitMine’s stock has traded heavily as investors assess the premium they are willing to pay for exposure to its ETH holdings and staking revenue.
Topics
Cryptocurrency