Changpeng Zhao, Binance’s founder and the initiator of Giggle Academy, said large-scale adoption of AI agents will require cryptocurrency-based payment rails, arguing that autonomous software cannot rely on bank accounts or credit cards to make frequent, low-value transactions.
In an interview recorded on Oct. 24, 2026, Zhao said AI agents could eventually pay for individual tasks, data requests, content views and computing resources without human approval at every step. He framed crypto networks and stablecoins as the practical settlement layer for such activity, particularly where payments may amount to a fraction of a dollar or yen.
Zhao predicted that the combination of blockchain infrastructure and AI could increase the speed of asset circulation and transaction activity by between 1,000 and 1 million times within five years. That estimate is aspirational rather than a forecast backed by a published model, but it reflects a growing focus on machine-to-machine payments as a possible use case for public blockchains beyond human trading.
Ai agents need wallets rather than bank accounts
Zhao’s argument rests on the limits of the conventional financial system for autonomous programs. A software agent can be assigned a digital wallet and a spending limit, while opening and operating a bank account generally requires a legal identity, compliance checks and a human or corporate account holder.
Credit-card systems also tend to be poorly suited to continuous small settlements. Their fee structures, chargeback mechanisms and authorization processes were built around consumer purchases rather than an AI model paying a few cents for an API call, a dataset query or a single article.
Zhao used a Japanese pricing example to illustrate the difference. Rather than paying 20,000 yen for a monthly content bundle, he said an agent could pay one yen to access a specific item. Such a model would depend on low-cost settlement, rapid confirmation and automated controls over the agent’s budget.
That scenario would favor stablecoins and networks designed for inexpensive transfers, though it also raises operational questions around wallet security, spending permissions, fraud prevention and the legal responsibility for an agent’s actions. Payments may be automated, but the organizations deploying the software would remain responsible for how funds are used.
A technical argument shaped by early trading systems
Zhao tied his interest in fast settlement systems to his first full-time software role in Tokyo, where he worked on order-routing and matching software connected to the Tokyo Stock Exchange. He said the job involved C++ programming, manual memory management and low-latency system design.
Those engineering concerns later influenced Binance’s early operating approach, according to Zhao. Crypto exchanges must handle surges in order flow, maintain matching engines during extreme price movements and process large volumes of account activity with minimal delay. The same focus on throughput now informs Zhao’s view that blockchains could become infrastructure for automated commerce.
Zhao said he lived and worked in Japan for roughly six or seven years. After Binance launched in Shanghai in July 2017, he said the company moved its team of about 30 people to Tokyo’s Meguro district later that year.
Japan was a prominent cryptocurrency market at the time. Zhao pointed to the country’s April 2017 legal recognition of bitcoin as a payment method and the start of exchange licensing later that year. He said around 16 platforms were approved in an initial group during September and October 2017.
He also said overseas customers represented about 70% to 80% of Binance’s potential trading volume in its early phase. Binance reached the top position for single-day exchange volume on Dec. 17, 2017, less than five months after its launch, Zhao said.
Japan’s policy choices remain part of Zhao’s case
Zhao described Japan’s token-screening rules as a commercial constraint for local venues during the 2017 cycle. He said Japanese exchanges commonly offered about 10 to 20 tokens, while Binance built its early business partly around a wider range of listed assets.
He said Binance spent $6 million in 2017 to compensate users affected during a period of regulatory disruption, an amount he described as equal to 40% of the company’s treasury reserves at the time. He also connected a $1 million donation following the July 2018 western Japan floods to the formation of Binance Charity. Zhao said 81 people died in the disaster.
His comments arrived as Japan considers changes to the tax treatment of crypto-asset gains. Zhao referred to a proposal that would reduce the top rate from as high as 55% to about 20%, bringing it closer to the treatment of stock gains. A lower rate could alter domestic trading incentives and reduce a long-standing complaint among Japanese crypto users, though any effect would depend on the final legislation and implementation rules.
Zhao said Binance returned to Japan in 2022 through an acquisition route that resulted in local licensing. He also referred to cooperation with SoftBank and its founder Masayoshi Son, without providing further commercial details.
Legal case and education project shape Zhao’s post-binance agenda
Discussing U.S. enforcement action, Zhao said he pleaded guilty to a single violation of the Bank Secrecy Act and served four months in the United States. He said the case did not include fraud allegations or claims of user-asset losses.
The former Binance executive compared the matter with a separate compliance-related penalty against Bank of America, while acknowledging that the comparison concerned the scale of penalties rather than the nature of the charges.
Zhao now says he funds Giggle Academy at roughly $5 million annually. The education project employs about 50 people and uses AI tools and game-like lessons to provide free learning through an app.
He cited more than 700 million illiterate adults worldwide and more than 500 million school-age children outside formal education as the scale of the problem the project aims to address. Zhao said registered students stood at about 100,000 in early 2026 and active learners later exceeded 2 million.
He said he could personally finance Giggle Academy for 20 years. Its longer-term model could ask future beneficiaries to voluntarily contribute 5% of their income to support the program, creating a funding mechanism linked to improved earning potential rather than upfront tuition.
Zhao’s machine-payment thesis places crypto less as a speculative product and more as an automated settlement tool. Whether AI agents generate payment volumes large enough to justify his predictions will depend on whether developers can build systems that are cheap, secure and trusted enough for software to transact on its own.
Explore how AI and blockchain integration could power crypto-native micro-payments and autonomous trading agents in tomorrow’s financial systems.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
