[DigitalToday intern reporter Seung-a Yoo] In the crypto asset management market, an approach is emerging that combines human traders' judgement with artificial intelligence (AI) analysis rather than leaving trading entirely to AI. The structure aims to reduce risks tied to automation and leveraged trading by excluding leverage, focusing on spot trading and spreading funds across multiple traders.

Cryptopolitan, a blockchain media outlet, reported on Oct. 5 (local time) that because crypto markets operate 24 hours a day, prices can move while users are working or sleeping. Automation can be an alternative for investors who find it difficult to constantly check charts. But leaving every decision to a trading bot has limits in responding to factors that are hard to explain with past data alone, such as news, liquidity, investor sentiment and unexpected variables.

Some platforms are choosing to divide roles between humans and AI rather than decide which is better. Skilled traders can interpret abnormal market conditions and revisit existing assumptions. AI can analyse large volumes of information that one person cannot handle at the same time.

Arbitflow was introduced as an example that applies this structure. Professional traders make actual trading decisions, while its in-house AI supports market analysis, news monitoring, performance evaluation and identification of risk signals. The approach is that AI does not replace people but provides additional information needed for analysis.

In trading methods, it promotes spot trading that does not use leverage. Leverage allows larger positions with less capital but also magnifies losses. If prices move unfavourably, there is also the risk that an exchange may forcibly liquidate positions. Spot trading, by contrast, involves buying and selling the underlying crypto assets directly without borrowing additional funds, so forced-liquidation risk from leverage does not occur.

That does not mean spot trading is safe. Crypto prices such as bitcoin and ethereum can fall sharply, and losses can occur if the entry timing is wrong or professional traders make the wrong judgement. In a prolonged bear market, the overall portfolio may also be affected. Eliminating leverage reduces certain risks but does not remove overall investment risk.

AI's role is also focused on decision support rather than fully automated trading. In a global market that moves 24 hours a day, price changes, economic announcements, regulatory shifts, exchange moves and crypto-related news can occur at the same time. AI can be used to detect abnormal price moves and compare them with past trends, analyse related news and identify changes in a trader's recent performance.

Such analysis does not necessarily lead to automated trading. Under the structure, professional traders who understand the strategy use AI information as a reference and make the final decisions. Arbitflow says in public materials that it began market research and strategy development in 2023, and that its in-house AI analyses not only market information but also traders' behaviour.

Another feature is that funds can be spread across multiple traders. Arbitflow said it conducts identity verification, reviews professional experience and runs simulated skill tests during trader selection, and then continuously monitors performance. Users can check traders' past performance, and they can also view most trading activity through a "live trading" function. The minimum entry amount is $25.

Past performance does not guarantee future returns. Users need to check how performance indicators are calculated, what the fees and fund-management methods are, and how the platform responds during periods when losses occur. Even if funds are spread across multiple traders, losses cannot be avoided if the overall market falls.

In the end, the core of this model is not about choosing between humans and machines. AI has strengths in quickly analysing and monitoring large-scale information, while professional traders can add context and experience in market conditions that are hard to judge by models alone. The approach is being presented as an alternative to fully automated trading or leveraged trading, using spot trading as a base, employing AI as an analytical tool and leaving final trading decisions to humans.