Binance co-founder He Yi said the exchange is taking an “All in AI” approach and wants financial products to become as easy to use as short-video apps, according to remarks she made during a Chinese-language AMA at Binance’s smart product launch event on Oct. 6, 2026.
During the session, He Yi spoke about Binance’s AI strategy, financial inclusion, product interaction, user data, the platform’s business model, and her recent personal situation. The discussion was later compiled by WuBlockchain from Binance Square content.
From trading venue to broader financial infrastructure
He Yi said Binance wants to become financial infrastructure for the future. She described the current moment as one where AI has already arrived, but many products and user experiences still lag behind. Using a metaphor, she said it feels as if electricity and the light bulb have already been invented, while the products people use are still in the era of candle factories. Looking at candlestick charts every day only reinforces that impression, she said.
For that reason, Binance has made it clear this year that it wants to use both blockchain and AI to drive product change across the platform. Trading may be the business most users focus on, but wealth management is closer to daily life and payments are the more routine use case, she said. Binance’s trading products are no longer limited to crypto assets, according to He Yi, and now also include precious metals and stock-related traditional assets, while the company is also building out savings, wealth management, and payments.
She added that investing always carries the possibility of gains and losses, and users need to manage risk. When users place assets on the platform, that represents trust, which in turn means Binance has to spend more effort, money, and resources on security every day.
He Yi also described Binance Square as the company’s social product. In crypto, she said, people can feel isolated, and the industry is often misunderstood by outsiders as something tied to gray-market activity or simply seen as strange. People who trade and invest regularly may even be viewed by relatives and friends as doing something improper even if they have made money. Binance wants users to find like-minded people on Square, she said, while also giving creators with real insight a place to be seen rather than buried on larger social media platforms.
In her telling, both blockchain and AI are only meaningful if they serve users. The technology should sit “under the table” and support the platform rather than dominate the user experience. She also said the old framing of competing with traditional finance and defeating Wall Street has given way, in her view, to a process of integration rather than a black-and-white replacement.
Why Binance is going all in on AI
He Yi said one of Binance’s core strategies is full commitment to AI. What she wants ordinary users to get from that, she said, is a form of equality — either financial equality or information equality.
Her description of the goal was straightforward: take mature analytical methods already used in finance, whether on-chain analysis or broader investment analysis, and present them in a simple, structured format. Even users with no investing background should be able to rely on AI to understand the logic behind investing, learn what kinds of analysis exist, and see which methods fit their own preferences.
She said she often thinks from the user side. If she were an ordinary user, what would she actually need? If she were a user with AI coding ability and entrepreneurial skills, what could Binance offer? She said she may not be a professional product manager or engineer, but sees herself as good at looking at products from the user’s point of view.
In finance, she says, models are not enough
Asked why financial AI may require annual net investment of $100 million as a threshold, He Yi said it is true that people can already use agents and AI coding to build strategies and make them run. Some people may say they can write their own order placement or strategy tools with AI, she noted. But for Binance, the issue goes beyond models, computing power, and coding ability.
Security, data, and users are the deeper foundations, she said. Binance’s edge starts with security and then with the platform’s existing base-layer data, including historical and current data. In programming, there is a large amount of public code on GitHub. In trading and finance, however, there is nowhere near as much free and open data. Anyone who has trained AI models understands the value of clean data, she said.
Binance has built up nine years of data, according to He Yi, and wants to use layers of security protection to turn AI from something that feels distant into something ordinary people can actually use. Comparing startups with a company like Binance, she said she sees annual net spending of $100 million as a threshold in finance. She did not frame that number as something that determines the winner by itself, but as a real barrier. Small companies may move faster, she said, while large companies have more foundational data and stronger security accumulation. In an industry where poor security reduces everything else to zero, that matters.
Using AI to narrow the knowledge and access gap
He Yi said Binance is, at its core, trying to address “freedom of money.” In Chinese, she said, that means trying to solve the problem of financial equality. One layer of that is the right to be informed.
She said investment markets in Europe and the United States are more mature, and many users there are familiar with investing and buying stocks. But in many other countries, people still have little or no concept of stocks and investing. She also spoke about her own learning curve. Before the age of 20, she said, she had no clear understanding of investing. She started trading stocks in 2008 and only gradually began to understand it after that. Before then, she felt there was nowhere to really learn, and she was also following FOMO, much like many beginners entering crypto today.
Her hope is that AI can show new users that macro conditions and breaking news can affect markets and prices. For an individual project, whether a token or a stock, company earnings reports, public disclosures, revenue, and buybacks can all be tied to performance. Too many people buy a coin or a stock simply because a friend tells them it will definitely make money, she said. Binance wants to show users the drivers behind price moves.
She acknowledged the counterpoint that if everyone can use AI to learn these things, everyone could end up back at the same starting line. Even so, she said, the process still matters because it narrows the gap between people who have no concept of investing and those who do, gradually flattening that unevenness.
A second layer is lowering the barrier to participation. Traditional brokerages often have a high threshold for opening an account and becoming a user, she said. Binance wants blockchain and AI to let users participate in stock investing with as little as $5. AI can help not only with trading but also with wealth management, she added. For users in high-inflation countries, holding stablecoins and using savings products is also a form of financial equality. In her framing, that means equality of capability as well as equality of access.
For people who are already active traders, she said, the tools can still help them keep learning and use those tools better. Binance wants users to start with beginner-friendly products — clear, structured, plug-and-play functions that are easy to understand on sight — then move on to AI Pro and eventually try writing code themselves. He Yi said she may be a typical example of someone who is self-taught, and she wants users to have that same opportunity. In her words, that may be the direction Binance and the broader financial industry will take in the AI era.
Making finance feel more like using an iPad or scrolling short videos
He Yi said she believes less is more. The biggest problem with many products today is that they contain too much information, and new users often cannot make sense of what they see. That, she said, is why Binance made the home page fully customizable: keep what is useful to you, remove what is not.
As personal AI and personalized AI continue to evolve, she said, the home screen on one person’s phone may eventually look completely different from another’s. The next step is to let users hand more tasks over to AI through conversation, while still keeping the market moves and breaking news they care about.
She used two product examples to explain the standard she has in mind. Apple, she said, created a product of epochal importance because even a 3-year-old child knows how to open and use an iPad. The achievement was not just the device itself but the way the product lowered the threshold for use. TikTok and Douyin did something similar by making uploading, interaction, and watching extremely simple. He Yi said she once worked on mobile video products, and before this product form appeared, video services still had meaningful barriers for many users. Now, whether someone is 3 or 4 years old or in their 80s, the process of uploading and receiving information is smooth.
Products like that win for an obvious reason, she said: they fit human behavior and sharply cut the barrier to use. Her hope is that Binance can one day do the same for finance, trading, and everyday payments. That is also why the company has started working on voice input and voice output. Typing, she said, is itself a forced learned behavior, and she hopes a future can arrive where even typing and manual input are no longer necessary.
Minimalism is not just a home-page design issue, in her account. It is also about interaction design, and that process will continue to evolve. When asked about Binance’s biggest challenge, her answer was simple: talent.
AI as an execution layer for user instructions
He Yi said she is not a trader and does not think ordinary people can easily learn highly complex trading knowledge. So she keeps asking what Binance products should look like from the perspective of an ordinary user.
Her example was direct. If a user tells AI to buy a certain amount of BTC when the price reaches a certain level, the system should be able to execute that request without forcing the user to study a complicated trading interface. She said she herself is not very patient when trading. Sometimes she has placed an order that did not fill, and when the market rose, the original quoted price no longer worked, leaving her waiting. With an AI product, she said, a user should be able to hand over the requirements and constraints and let the system finish the task.
At this stage, she said, the goal is for AI to complete those actions smoothly: you provide the conditions, and it executes them. She added that this already exists, but she wants to improve it. Over time, she wants AI to judge a user’s preferences — whether that person is more of a long-term holder or an active trader, whether they prefer dollar-cost averaging or grid strategies — and then recommend structured investment approaches that only require confirmation. She said Binance also wants each person to have their own personal agent, “like Iron Man.”
User data, recommendations, and model training
On whether users are treated differently based on assets or VIP level, He Yi said she does not believe such differential treatment exists. The logic, she said, is to present existing knowledge in a structured way and then recommend content based on personal behavior preferences.
If she is a typical user who does not trade, she said, then pushing trading content to her every day would not be useful. She would care more about what happened in the market that day. That is also why she encourages users to customize the home page. If someone continues using the default template and does not interact much with the platform, Binance will not know that person’s behavior patterns or profile very well.
She said the platform may not currently go very deep into granular personal data. It works more by classifying users according to interests and behavior, similar to how TikTok labels users by interest categories. On the specific question of whether user data will be used for model training and whether that can be turned off, she said Binance may not yet have gone that deep in mining user data, but the suggestion will be considered.
How Binance makes money
He Yi said market swings mean some people make money and others lose it, and the past few years have felt like living in extremes every day. Profitable users send her direct messages saying “I love you,” she said, while people who lose money accuse her of “harvesting leeks” and assume the losses went into her pocket.
Her answer was that Binance follows a basic principle: it only earns the money it is supposed to earn. The company’s logic is to make money from fees, she said. First, it does not take the other side of users’ trades. Second, it does not take even a penny from users through hidden means. If there are problems, Binance acknowledges them, she said. It is not a perfect platform, but it will keep evolving. Users place money where they place their trust, she added.
She also urged users to be more cautious with their own money. Some people may see meme coins as a chance to take a gamble and turn a bicycle into a motorcycle, she said, but they should not spend every day looking for hidden messages in posts from her or CZ. The tokens that emerged in the last cycle did not rise because people parsed tweets word by word, she said, and that behavior has now become a profit model for some token issuers. Her advice was clear: do not get so wrapped up in FOMO, do not let others lead you by the nose, and keep your own independent judgment.
He Yi says she is pregnant again
At the end of the AMA, He Yi also addressed recent questions about her health. She said there have been special physical reasons recently that left her short on energy, and she has spent less time scrolling through posts. Even so, she said she still tries to read user feedback and requests about products and solve problems where she can.
She said many people had sent messages asking whether something was wrong with her health and telling her to take care of herself. Her response was that things were fine and that it was actually good news. At the Hong Kong event earlier this year, she said, she had in fact been pregnant, but because she was too tired at the time, she lost the baby. Now, she said, she is pregnant again.
Her workload has not been reduced, she added. She is still working every day in the small cubicle users have seen in photos and continues talking with the community, so she has indeed been busier and more tired. Still, she said her overall situation — whether at Binance, in life, or physically — is good.
The original compiled text also stated that the guest’s remarks do not represent WuBlockchain’s views and do not constitute investment advice, and that readers should follow local laws and regulations. It also noted that the audio transcription and organization were completed by GPT and may contain errors.