Artificial intelligence trading bots are emerging as a new trend in the cryptocurrency market. Interest among individual investors is rising as claims spread that automated bots using Anthropic's Claude have generated millions of dollars in profits.

But major AI companies do not officially support such use, and experts are urging caution, pointing to structural limitations.

On March 17, blockchain media outlet BeInCrypto reported that Haseeb Qureshi (하지브 쿠레시), a partner at crypto investment fund Dragonfly Capital, said the AI trading boom is largely exaggerated.

Qureshi said three conditions must be met for AI trading to truly succeed. Big Tech firms must develop dedicated trading models, individual investors must be able to compete with institutions, and AI must be able to generate profits consistently in open markets. He said all three conditions are difficult to meet in practice.

Qureshi first cited "liability risk" as the reason Big Tech is not actively developing AI models for crypto trading. If AI makes wrong trades that result in large losses or sends assets to the wrong address, it could become a legal liability issue rather than a technical one. He said major AI companies are therefore maintaining a cautious stance in the area.

Even if individuals build trading strategies using AI, they run into structural limits. Strategies based on AI models operate on open technology and are accessible to anyone, and institutional investors with faster infrastructure and more capital can ultimately run the same strategies at scale. "If there is money to be made with a base model, big quant funds would have already scaled it up by the thousands and be using it," Qureshi said, underscoring the limits of individual investors' competitiveness.

Expectations that autonomous AI agents can generate profits on their own are also seen as unrealistic. AI systems based on the same model produce similar strategies and ideas, making it difficult to secure differentiated competitiveness. Critics say they ultimately lack originality that comes from human experience and context.

For these reasons, analysts say AI trading is unlikely to establish itself as a sustainable profit model except in some cases. Markets are already seeing competition among institution-led ultra-fast algorithms and large pools of capital, making it virtually impossible for individuals to compete on the same terms.

Overall, the AI-based trading boom shows that even if some profit cases exist, it is difficult for individual investors to maintain a long-term edge in a market that structurally favors institutions. The industry has said that the rule that "the house wins" applies in AI trading as well.