CZ argues AI agents will bring crypto-powered finance to billions worldwide

Changpeng Zhao, the founder of Binance, told Galaxy Research that AI may pose an existential risk to humanity, but not to crypto. His argument is that digital assets could become the financial rail for autonomous AI systems before traditional banks are ready for machine-driven payments.

The Galaxy Research interview, released on June 18, covered Bitcoin’s market cycle, Satoshi Nakamoto’s crypto holdings, Hyperliquid, and the future of AI agents. Zhao’s strongest point was that autonomous software will likely adopt crypto rails before legacy finance adapts to non-human economic activity.

CZ says AI can extinguish civilization. Crypto can’t.

The founder of Binance sits down with Alex to talk Bitcoin cycles, Satoshi’s coins, Hyperliquid, and why AI agents will use crypto before banks even realize what happened. pic.twitter.com/jnBLNxPnBQ

— Galaxy Research (@glxyresearch) June 18, 2026

That argument also has a financial-inclusion angle. The World Bank’s Global Findex Database 2025 shows that global account ownership has reached record levels, but about 1.3 billion adults still remain outside the formal financial system. Permissionless, internet-native payment systems could, in theory, serve both autonomous AI agents and people excluded from traditional banking.

CZ says AI is a human risk, not a crypto risk

Zhao’s position separates the risk of AI from the opportunity for crypto.

AI may create risks for jobs, security, and human control, but Zhao does not see it as a threat to digital assets. Instead, he argues that AI agents will need payment systems that are open, programmable, and available across borders.

That is where crypto enters the picture. Autonomous software cannot easily open a bank account, complete corporate onboarding, or wait through traditional compliance workflows. A crypto wallet, by contrast, can be created quickly and used to send value globally.

The point is not that crypto solves every AI risk. It is that AI agents may need a payment layer that moves at software speed, and crypto already offers that structure.

AI agents could make crypto payments more useful

Zhao has made this argument before. At the World Economic Forum in Davos, he said, “The native currency for AI agents is going to be crypto,” according to DL News.

Circle CEO Jeremy Allaire also said at Davos that billions of AI agents could be conducting economic transactions within the next three to five years. If that happens, payments between software agents could become a major use case for stablecoins and blockchain rails.

The logic is simple. AI agents may need to pay for data, API calls, compute resources, storage, and other digital services without human involvement. Traditional banking systems were designed around people and companies. Crypto wallets were designed for internet-native value transfer.

Analysts at Andreessen Horowitz have described a similar future, where AI agents pay each other instantly for data, GPU time, or API calls without invoices, reconciliation, or batching. If that model scales, the same rails could also support human users in markets where banking access remains limited.

The unbanked argument gives the idea a wider frame

The financial-inclusion angle is what makes Zhao’s argument larger than a market-cycle call.

The World Bank’s Global Findex data shows that hundreds of millions of adults still lack access to formal financial accounts, even as mobile phone ownership and digital payments continue to expand. That gap is especially important in emerging markets, where people may have internet access but not full access to dollar-based banking, cross-border payment tools, or affordable credit.

Crypto does not automatically solve that problem. Users still face volatility, regulation, fraud risks, poor user experience, and unreliable on-ramps.

But if AI-driven commerce pushes more activity onto blockchain rails, the same infrastructure could become more useful for people outside the banking system. Stablecoins, wallets, and low-cost settlement networks could support remittances, savings, small loans, and cross-border payments without requiring every user to pass through a traditional bank.

That is the financial-inclusion case behind Zhao’s AI argument. Crypto is not only a speculative asset class. It could become a shared payment layer for both autonomous software and underserved users.

Crypto winter has not changed CZ’s long-term view

Zhao remains bullish on Bitcoin and crypto, even while stepping back from earlier timing calls.

In January, he told CNBC that Bitcoin could break its traditional four-year cycle and move into a supercycle because of more favorable crypto policy in the United States and other countries.

More recently, he has acknowledged that the market has not delivered the supercycle many expected. But his long-term view has not changed. His argument is that crypto continues to grow through new use cases, even when prices weaken.

That view now extends beyond trading. Through YZi Labs, Zhao remains focused on blockchain, AI, and biotech investments. He has also said he prefers investing in “AI shovels,” including infrastructure such as data centers, power, and computing hardware, rather than chasing every AI application directly, according to Yahoo Finance.

His post-Binance role also shapes the argument. Zhao no longer runs the exchange after pleading guilty in 2023 to violating the Bank Secrecy Act by causing Binance to fail to maintain an effective anti-money laundering program. Binance agreed to pay $4.3 billion in penalties, while Zhao resigned as CEO, according to the U.S. Department of Justice.

The $20 trillion forecast depends on real adoption

The AI-crypto link has already attracted big projections.

Bitwise previously estimated that AI and crypto together could add $20 trillion to global GDP by 2030. DL News also reported that investors put more than $565 million into startups working across AI and crypto in 2025, citing DefiLlama data.

Those numbers show investor interest, but they do not prove adoption. The real test is whether AI agents actually use crypto for payments, settlement, identity, and coordination at scale.

If traditional payment processors adapt quickly, regulated fintech rails could capture much of the market. If they move slowly, crypto networks may have an opening because they already support programmable, global, wallet-based transactions.

That is Zhao’s bet. AI may be dangerous for humanity, but it could also create one of crypto’s most practical use cases. If autonomous agents start using digital assets for everyday transactions, the same rails could eventually serve millions of people still excluded from traditional finance.