In April, top Chinese retail stock broker East Money organized a virtual trading competition, asking its users to deploy an OpenClaw artificial intelligence agent to trade a portfolio of stocks with a RMB 500 (USD 73.9) prize.

The competition took place in a simulated environment and did not involve real money. But trading stocks using AI bots has already become a reality for retail investors.

Investors can use AI by connecting a large language model (LLM) to the stock broker’s application programming interface (API), inputting trading strategies and then letting AI do the work. Major Chinese brokerages including East Money have rolled out functions that allow retail users to connect OpenClaw with their databases to receive summaries and stock recommendations.

In China brokerages still ban employees from using AI tools for professional trading, citing data privacy concerns and execution risks.

David Friedland, Asia Pacific managing director for Interactive Brokers (IBKR), which operates the largest electronic trading platform in the US, noted that it remains challenging for platforms like IBKR to track a client using an AI agent like OpenClaw to write a model that trades on their behalf.

“We can measure the speed and the duration of the orders coming in, but that can be a computer program too, so it’s really hard for us to say what people are using,” he told Nikkei Asia. “But there is no doubt we will see more people using AI models to write their own codes for trading.”

In Hong Kong, where investors can easily access a range of investment instruments from stocks to derivatives across most international markets, securities brokers are racing to add AI capabilities to their trading platforms, though they are much more cautious when it comes to fully automated executions due to regulatory concerns.

Based on DeepSeek’s LLM, Futu Holdings, a retail market-focused stock brokerage headquartered in Hong Kong that owns Futubull and Moomoo, launched its AI assistant “skills” in March. Clients can install the assistant on open-source platforms such as OpenClaw, Claude Code, and Cursor to execute trading strategies, though they must provide a trading password as final confirmation.

Futu has also launched AI services that allow users to generate quantitative trading strategies using natural language, without the need to know coding.

“AI’s impact on trading may be subtle in the near term, but its long-term significance is undeniable. We’ve seen a positive correlation between AI usage and trading activity,” said Vincent Yao, head of the AI growth center at Futu Holdings.

The firm tracks AI usage through user interaction with AI-generated content such as financial analysis.

Yao admitted that certain features of its chatbot NiuNiu AI are limited due to regulatory requirements despite being built as an AI agent. Currently, the AI bot is only implemented in Q&A settings.

The world’s leading AI companies have set their eyes on financial services in a bid to grow their enterprise client base. Last week, Anthropic launched 10 new AI agents aimed at what it calls “the most time-consuming work” for the financial services sector. The agents will be able to help build pitchbooks, screen know-your-customer files, and closing books.

In financial markets, the use of AI is quickly blurring the line between data analysis and investment advice. In most regulated markets including Hong Kong, providing such advice requires a license and implies a higher level of responsibility towards clients.

Hong Kong does not currently have AI-specific regulations for financial institutions but relies on existing regulatory frameworks. In a circular published by the Securities and Futures Commission in November 2024, the securities watchdog states that using AI to provide investment recommendations, advice or research as “high risk use cases” and urge licensed companies to provide risk mitigation measures.

“AI is changing so fast that the regulation is going to be catching up for a while,” said IBKR’s Friedland.

“All of a sudden, with AI, everyone can look like the world’s greatest options trader,” Friedland said of the risks involved in AI trading, “but the reality is they need training and experience to understand the risk behind their trading.”

Like Hong Kong, other Asian economies are mostly applying existing securities rules to AI systems. In November 2025, the Monetary Authority of Singapore released a consultation paper, “Guidelines on Artificial Intelligence Risk Management,” requiring institutions to assess how risky each AI use case is before rolling it out and making sure any remaining risks stay within the firm’s tolerance levels before deployment. The agency also launched a toolkit in March to help financial institutions manage risks in AI.

“Regulators are concerned that AI may turn flawed signals into real trading, advice or suitability outcome at speed and scale,” said Tommy Liu, global regulatory partner at law firm Hogan Lovells.

China integrates financial AI governance into its strict AI governance framework through a stack of vertical rules, and regulators there could deem AI-driven trading recommendations problematic. A securities company in Shanghai was fined for RMB 2 million (USD 295,460.3) by regulators in 2025 after it failed to disclose algorithmic limitations in its AI-generated investment recommendations.

In practice, the problem is not usually an absence of law but the evidence that controls work. Liu noted that companies need “explainable governance, documented testing, human escalation, real-time monitoring, robust data controls and vendor accountability.”

So far, AI alone has not delivered the promise of eye-watering returns for investors.

At the end of 2025, finance-focused AI research lab Nof1 launched an AI trading competition among eight top AI models. They were handed USD 10,000 to trade US tech stocks for two weeks and were given free rein to take risks and use leverage. Only six out of 32 attempts across these models delivered positive results, with one Grok 4.20 model being the winner with USD 34.59% return.

“It’s only a matter of time before people pile into AI bots trying to make money,” Friedland said.

“The next stage of competition among brokerages will hinge on how effectively firms can turn information into intelligence, shifting focus from transaction speed to insight creation,” said Jingxin Du, TigerAI product lead of Tiger Brokers, a Chinese brokerage firm headquartered in Singapore that focuses on the Hong Kong and US markets.

According to Du, TigerAI’s cumulative global interactions surpassed 10 million by the end of March, a 500% jump since its launch one year ago, while its total user base grew by 148% over that period.

Daniel Tse, managing director at Futu, said that to him investment advisory could be an area where AI can play a bigger role in the future, though service providers need to overcome significant hurdles.

“It’s not just about regulatory compliance. It’s also about how you train AI to ensure its recommendations are unbiased. That’s a core challenge the entire industry is still trying to solve,” Tse said.

This article first appeared on Nikkei Asia. It has been republished here as part of 36Kr’s ongoing partnership with Nikkei.

Note: RMB figures are converted to USD at rates of RMB 6.77 = USD 1 based on estimates as of June 15, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates.