AI Stock Trading Software Pitfall Guide: Legal Risks for Programmers Based on Real Cases

As the AI wave sweeps through the financial sector, countless programmers and entrepreneurs are passionately diving into “hot” tracks like quantitative trading and intelligent investing. But what you might not know is that a seemingly innovative stock trading software has already caused many to fall into legal trouble. A real case that occurred in Shanghai has sounded an alarm for the entire industry—this time, the offenders include not only company owners but also technical developers who are not immune to legal sanctions. What kind of legal traps are hidden behind this?

Lesson worth 30 million: How S Company Went Down the Illegal Path

The involved S Company appears on the surface to be a tech enterprise, but what kind of business is it actually doing? The company, without approval from financial regulatory authorities, sold two types of products through an online platform called “Xundong Quant”: one is a “range arbitrage” quantitative trading strategy, and the other is a “DIY stock trading robot” software tool.

Sounds very “high-tech,” right? But the real issue lies in the underlying operational logic. S Company also provided what is called “securities consulting services”—recommending specific stocks to buy or sell, and when to do so. In other words, users not only bought software but also purchased “investment advice.” With this business model, S Company accumulated profits of over 30 million yuan.

The final penalty was severe: the company’s actual controller, Zhong, was sentenced to seven years and nine months in prison. Not only Zhong, but also the company’s technical director, Kong, was penalized for participating in the development of this system. Even more outrageous, the company purchased hacking software from a hacker named Han to bypass the technical protections of the market terminal software (Tongdaxin). Han was sentenced to three years for providing intrusion programs.

This is not an isolated incident. It reflects a widespread problem in the industry: too many AI entrepreneurs and programmers are blurring the legal boundaries or even unknowingly crossing red lines.

Three Dimensions to Recognize the Bottom Line: What Kind of Stock Trading Software is Illegal

So the question is—why is it that some companies developing stock trading software are illegal while others are not? Where exactly did S Company go wrong?

Function Output: The Boundary Between “Providing Data” and “Providing Advice”

In S Company’s 8,800 yuan membership package, what does the “range arbitrage” feature do? Simply put, the software analyzes the stock’s recent price movements over the past few days and then tells users “Now should buy this stock” or “Now should sell that stock.” This is no longer just a “data analysis tool” but directly outputs “investment advice.”

According to regulatory rules, what should a legitimate information tool do? It can provide objective facts like capital flow data, sector anomalies, historical transaction data, etc. But it absolutely cannot say “Buying A-shares has a high probability of profit” or “B-shares are about to rise”—these fall under “investment advice,” which can only be provided by licensed investment consulting institutions.

S Company not only provided advice but also had customer service staff perform “further analysis and recommendations” based on experience. This effectively turned a tool company into an investment consulting firm, which requires special licensing and regulation. That’s why platforms like Tonghuashun and Eastmoney offering quantitative tools are legal, but S Company’s “DIY stock trading robot” is illegal—because the former only provides data and tools, while the latter offers investment decision-making.

Business Logic: Who Pays and Why

Look at S Company’s charging model. Why do users pay 8,800 or even 28,800 yuan? Not just for a software, but because of a promise: “This system can help me make money.” Essentially, users are paying for “profitability.”

This is a typical business model of investment consulting and trading facilitation services, not the pricing logic of a technical tool. If a software only charges a one-time fee or a low-cost data subscription, and its main value is providing information itself, it’s closer to a “technology service.” But if it adopts tiered memberships and prices based on “ability to make money,” it’s a disguised sale of investment services.

Operation Loop: Who Makes Decisions for Investors

S Company’s high-end 28,800 yuan membership feature works like this: users set certain conditions (e.g., “when this stock rises more than 5%”), and then the software automatically completes the entire process from analysis, decision-making, to order placement. Investors don’t need to do anything; they just wait for the software to trade for them.

This is the core problem. The normal investment process is: investors receive information → analyze themselves → decide → place orders. But S’s model becomes: software analyzes on behalf of investors → makes decisions for them → executes trades. This effectively makes S Company act as a “trade executor,” a role that should be performed only by licensed securities firms.

The Black-Hand of Technology: Turning Tools into Crime Devices

There’s an even more sinister aspect to this case. S Company’s ability to automate trading so efficiently relies on illegally cracking security measures. They purchased an “外挂” (hacking/cheating program) from a programmer named Han and embedded it into their system. What does this外挂 do? It bypasses Tongdaxin’s technical protections, directly intruding into the broker’s trading system illegally.

In other words, the normal trading path is investor → broker. But with this “black tech,” S Company created a side route: investor → S Company → (illegally accessed) → broker. Then S Company charged users a “connection service fee”—over 3 million yuan—just like that, flowing into their pockets.

This makes the criminal nature of the case even more serious. It involves not only illegal securities business operation but also:

• Invasion of computer systems: Han was sentenced to three years for providing intrusion programs.
• Copyright infringement: Unauthorized use of Tongdaxin’s code and interfaces.
• Illegal business operation: The actual controllers and technical leaders of S Company were convicted on multiple charges.

In short, what appears to be a “technological innovation” is actually a combination of multiple illegal acts.

A Programmer’s Must-Read: Red Lines You Must Not Cross

What does this case mean for programmers working in fintech? If you are developing or considering entering this field, you must understand several bottom lines:

First, do not cross the line between “providing tools” and “providing advice.”

You can develop software that displays stock data, alerts sector anomalies, calculates technical indicators, etc. But you cannot write code that automatically outputs commands like “recommend buying.” Once your code’s output is “advice” rather than “data,” you shift from being a technical developer to an investment advisor, which requires a financial license.

Second, do not use technical means to evade regulation.

Hacking interfaces, embedding外挂, bypassing protections—these may be called “highly skilled” in cybersecurity, but in finance, they are serious crimes. Programmers involved face not only company risks but also personal criminal liability. Han is a living example.

Third, the business model itself can reveal issues.

If your product charges based on “whether it can help users make money,” that business logic indicates you are selling investment services rather than just a technical tool. No matter how clever the legal language, this point cannot be hidden. Regulators and judges will look at this.

Fourth, have a clear role definition.

In a compliant financial system, roles are clearly divided: exchanges set trading rules, brokers provide trading channels, investment advisory firms offer advice, and tech companies supply tools. What role does your company play? If the positioning is unclear, it’s easy to slip into violations.

What products are compliant? A financial data terminal that only displays information and performs basic calculations, without outputting investment advice; or a quantitative backtesting tool that helps users design and test strategies but does not execute trades automatically; or an API service allowing licensed institutions to access data—these are all legal. The key is to clarify whether you provide “tools” or “services.”

Summary: Compliance is Not Restriction, But Protection

The story of S Company has become a textbook case. From this, we see that the legal attitude towards stock trading software is not to stifle fintech innovation but to define clear boundaries: you can develop technology, but you cannot cross into operating financial services.

For AI entrepreneurs and programmers, the smartest approach is not to try to skirt the rules but to conduct compliance assessments during product design. Consult legal experts, clarify whether your product ultimately outputs “data” or “advice,” whether your charging logic implies investment promises, and whether your technical solutions bypass necessary regulatory steps. This is not trouble but protection—avoiding legal risks, safeguarding users, protecting yourself, and ensuring the industry’s sustainable development.

As fintech and AI become more prevalent, such cases may recur. But industry practitioners should learn from S Company’s lesson: true innovation should be based on rule-abiding development, not rule-bending.