Investment decisions are increasingly shaped by what people see others do online, according to Vugar Usi, chief executive of crypto exchange MEXC. Speaking to BeInCrypto at TOKEN2049, Usi described a “screen-optimised life” in which people observe someone else’s successful investment and try to replicate it.
Artificial intelligence adds another layer of advice, with growing ability to act on an investor’s behalf. On September 28, MEXC launched a tool connecting users’ own AI agents to the exchange, allowing them to prepare and execute trades, with confirmation required for key actions.
Social trading as an identity signal
Usi sees social trading becoming “a kind of identity signal,” where an investment expresses who someone admires or how early they recognised an opportunity. “Sometimes you want to be able to say: I invested in Amazon ten years ago. I was one of the first to discover a certain opportunity,” he said.
Research supports taking those motives seriously. A FINRA Foundation study published in April, using its 2024 investor survey, found that social activity motivated 59% of investors using social media, compared with 11% of non-users. The same research found a gap between confidence and tested investment knowledge among these users.
Usi also noted that enthusiasm for AI companies helped drive trading on MEXC. “That’s the topic everyone is talking about, and it fuelled the trade,” he said. Following a popular theme makes opportunities easier to discover, but deciding whether the price makes sense still requires separate work.
AI and the quant advantage
Usi believes AI can help individuals use some of the analytical logic long available to professional trading firms. “Quant firms have been doing algorithmic trading for years. What AI is doing is helping individual users benefit from that kind of logic,” he said.
Professional trading also depends on tested strategies and reliable data. Giving someone an AI assistant does not establish that they can reproduce a quantitative fund’s results.
MEXC’s October 15 survey of 3,708 crypto traders found that 84.8% had used AI for research or trading. Among AI users, 65.9% acted on its recommendations at least sometimes. Those are findings from the company’s sample, and reported improvements reflect users’ impressions rather than independently measured returns.
Usi expects the next step to involve greater delegation. “Soon I believe we’ll see a lot of autonomous agents that can rebalance your portfolio and act on opportunities,” he said. That shifts the investor’s task towards evaluating a system whose individual decisions they may struggle to inspect.
Overnight trading and human oversight
The attraction becomes clearer as trading hours expand. “The saying goes that money never sleeps, and neither do exchanges. But humans do sleep,” Usi said. Nasdaq plans to introduce 23-hour trading, five days a week, on December 6, extending the window in which investors can respond to events ahead of 2027.
Usi imagines software managing positions during those hours. “Maybe in the future, instead of relying on liquidation and stop losses, AI agents will be able to make better decisions before you reach those danger zones,” he said. That remains a proposed benefit; the interview supplied no performance evidence showing that AI prevents forced closures of leveraged trades.
For now, Usi describes a human approval step. “They can give you an idea of what to do, and all you have to do is confirm the trade before it’s executed,” he said. A confirmation button leaves the investor with a consequential job: understanding what they are approving.
Usi’s broader position is that safeguards should stay in place. “That’s what we’re aiming for with AI: you should be cautious about the dangers and limitations, and you shouldn’t waive those limitations, especially in finance,” he said. The Financial Stability Board’s October 15 guidance calls for human oversight suited to an AI system’s autonomy and risk.
How much would users hand over?
MEXC’s survey suggests users draw boundaries. In a scenario that prohibited transfers and withdrawals, 54% would let AI trade with no more than 20% of their account assets. Only 6.5% would authorise more than half.
Usi also recognises the commercial importance of the customer relationship. “Either you have the customer and you can bring all the services to them, or you don’t,” he said. That relationship becomes more consequential when a platform helps shape an investment idea and provides the machinery to act on it.
By 2027, more investors may spend their time supervising decisions they once made themselves. Their control will depend on whether they understand the authority they have delegated—and can withdraw it when the strategy no longer fits their needs.
