The advent of artificial intelligence heralded a new era in both data processing and the utter diminishment of human critical thinking and information analysis. The ability of AI to process vast amounts of information makes it an ideal instrument for the identification of possibly beneficial grains of data that can be used to derive insights and trends. However, the forecasting abilities of AI are extremely limited, since they rely on the information that the machine is provided with. Considering that the AI is devoid of the ability to apply critical thinking and distinguish truth and facts from garbage data, the reliability of the derivations it produces is minimal.

The low reliability of AI as a trend identifier has not stopped traders from applying it well beyond its initial domain of acting as an agent for automating trading strategies. While traders had previously only delegated strategy application to AI, many of them are now relying on the conclusions and recommendations of AI for formulating not only their trading strategies, but also their expectations from market behavior.

A major illustration of the erroneous perception that such generative-AI tools as ChatGPT can be used for predicting the possible price movement of coins can be found in a recent report released by the Bybit exchange. The report urges users to refrain from such a practice, as statistical research into the performance of users relying on the chat bot revealed that it is very far from ideal. AI-powered trade deals were consistently found to be unprofitable. Of the 1,100 trades made based on recommendations from ChatGPT, all 100% resulted in losses for users at a total of $2 million, on average $2,000 per user.

Risks of Using AI in Trading

The promise of improving efficiency and profitability through the analysis of vast amounts of data has driven the use of AI into the trading arena in recent years. And though the benefits of AI are many and span a number of industries, their use in trading is still questionable and should be done with extreme caution. Among the dangers that the use of generative AI in trading can entail are:

Algorithmic Price Manipulation

Basically put, generative AI tools can be used to artificially manipulate stock prices. The ability of such tools to produce vast amounts of data on demand based on individual whims and prompts makes it hazardous for other market players to trust the market data they are viewing. The error of false prediction trails the benefits of using AI for deriving insights from available market statistics.

The increasing adoption of AI and the swelling competition among generative AI developers is resulting in the advent of malicious players who are pursuing personal agendas in the use of AI. By manipulating market data, such players can trick inexperienced traders and make them believe in false information, thus convincing them to make false investment decisions and incur losses as a result.

Historical Data Caveats

The biggest problem with data analysis using AI is that the machine cannot discern truth from false. This applies to historical data, which may not include some vital inputs that are essential for deriving correct and accurate predictions by AI. Such a pitfall is especially acutely felt on the futures market, where modeling strategies on the basis of historical data is key. With a lack of input variables, the AI makes mistakes and reliance on such outputs can result in losses. The data produced may not always reflect not only reality, but even a shadow thereof. The reason is the huge amount of garbage and false data that is also processed by the AI, effectively diluting valuable statistics.

Lack of Transparency

Though already widely spread, generative AI systems remain poorly accessible and understandable to the average user. The marketing ploys utilized for their spread omitted the many caveats and technical specifications that AI entails. This has resulted in a ‘black box’ problem, meaning that users do not actually understand how AI works and what it can actually be used for. The result is a lack of transparency, compromised personal data security, and lack of an even playing field among users of AI.

Cascade Effects

As the use of AI in trading increases, overreliance on such systems can create the risk of a cascade effect. This means that a single failure in one of the major AI system providers will spill over in a domino effect onto others, resulting in the collapse of the market as a whole. This systemic risk poses an immense threat to the entire global financial ecosystem, driving regulatory authorities to closely monitor both the developers and users of generative AI tools. Traders must be educated about the risks they are facing by using AI and measures should be developed and implemented to mitigate the possible consequences.

Disruption

The growing use of AI among traders is placing established trading strategies and patterns under threat. With AI being used for both reacting to market movements and generating them, the level of flux and volatility in markets is expected to increase, resulting in sudden shifts that can result in considerable consequences. Though it is impossible to neglect the ability of AI to identify new and sophisticated approaches to trading and innovative strategies, the risks involved are yet to be evaluated.

Disadvantages of Use of AI

The stock market is the biggest arena for the use of generative AI. With trillions of dollars in volume, the market is highly susceptible to the impact excessive use of AI can entail. The following is a list of the main disadvantages of the use of AI in stock trading:

  • No human judgment – the lack of emotional intelligence in AI can result in the oversight of important factors affecting stock pricing, such as quality and long-term potential.
  • Excessive dependence – historical data cannot act as the sole gauge of stock indices, while AI relies on such data for making predictions, neglecting sudden external and internal factors.
  • Bias – data bias is a constant for AI, since the machine relies only on the data that was used to train it, neglecting other, potentially impactful factors, resulting in a perpetuating chain of unfairness towards some market participants.
  • Amplifying effect – the use of AI can result in increased, artificially provoked, and unjustified price volatility through excessive use of automated trading.
  • No adaptation – unlike a human, an AI has no creativity or imagination, has trouble adapting to sudden changes, and cannot react to external shifts, resulting in slower reaction timing, which leads to poor investment decisions.
  • Regulation – there is no regulation on the rampant use of AI, leaving oversight authorities far behind in reaction, opening the way for innovative types of market manipulation and fraud.
  • Security – all AI constructs can be easily breached by cyberattacks and hacking, leaving their systems and their operators’ funds vulnerable.
  • Economic impact – overreliance on AI can result in unfair job losses, leading to economic disparity and a serious widening of the gap between social layers.

Conclusion

Artificial Intelligence may be a useful instrument, but its use and application in the trading and financial sectors require not only further extensive study, but also considerable regulation and caution. The lack of comprehensive research on the impact of such systems on trading is hampering their regulation, leaving the path wide open for a myriad of manipulators and speculators. With no safeguards to protect the market against AI, many investors and traders remain oblivious to the threats posed by AI-facilitated trading.