Lede
In 2025, the cryptocurrency market experienced a significant shift toward automated trading strategies. Data from HTX, a top-10 global exchange by volume and traffic formerly known as Huobi, indicates that traders increasingly relied on algorithmic tools to navigate market conditions throughout the year. The Seychelles-based platform reported that its grid trading volume rose by 97% year-over-year in 2025. This surge was accompanied by a doubling of the total capital allocated to grid-based strategies on the exchange.
The growth was particularly notable within specific market segments. Grid trading volume for stablecoin pairs saw an increase of 352% over the previous year. This figure significantly outpaced the 122% growth observed in grid trading for major cryptocurrencies. HTX’s position as one of the world’s 10 largest exchanges by liquidity and platform traffic provides a broad view of user behavior. The exchange’s recap highlights that the trend toward automation was not localized but represented a broader leaning toward algorithmic execution. By utilizing these strategies, traders have changed how they interact with spot markets, prioritizing automated orders that execute buy and sell instructions as the market moves within defined price ranges. These grid bots allow participants to set specific price ranges, facilitating the execution of orders as the market fluctuates back and forth, rather than relying on manual directional bets.
Context
Beyond traditional grid bots, the industry has seen an expansion in the use of more sophisticated AI-powered agents. Coinbase has been a primary driver of this development, expanding the use of AI agents across its suite of trading tools to enhance the user experience. This initiative dates back to at least August 2024, when Coinbase CEO Brian Armstrong confirmed the company had begun testing AI agents in real-world scenarios. In one early trial, a bot utilized cryptocurrency tokens to interact with another AI system to purchase training data, a process the CEO described as tokens buying tokens.
Following these initial tests, Coinbase introduced the Based Agent tool in October 2024. This tool was designed to allow users to create AI agents that are directly linked to cryptocurrency wallets. These agents are capable of performing various automated onchain activities, including trading, executing swaps, and managing staking operations. The development of such tools indicates a shift from fixed-rule execution, like standard grid trading, toward autonomous agents that can interpret natural language and operate independently on the blockchain.
The trend toward AI-managed finance is further supported by industry research. A survey conducted by CoinGecko revealed that approximately 36% of respondents expressed a willingness to allow AI agents to manage the majority of their cryptocurrency holdings. This sentiment underscores a growing comfort level among retail and institutional participants regarding the delegation of portfolio management to automated, intelligent systems that can operate 24/7 in the digital asset markets.
Impact
The introduction of new technological frameworks has further lowered the barriers for AI integration in the crypto sector. In October 2025, Coinbase launched Payments MCP, a tool specifically designed to enable AI agents to interact directly with onchain financial services. A key feature of this system is that it does not require the use of traditional API keys, which have historically been a requirement for external software to interface with exchange or wallet services. By allowing large language models to access wallets and onramps through natural language prompts, Payments MCP simplifies the process of automated financial interaction.
However, this increased accessibility brings new considerations for security and system architecture. Aaron Ratcliff, the attributions lead at the blockchain intelligence firm Merkle Science, has noted that granting AI agents access to crypto wallets introduces a specific set of challenges. According to Ratcliff, this shift adds a new layer of trust to systems that were fundamentally designed to be trustless. Because the AI agent acts as an intermediary with direct access to private funds, much of the security responsibility is shifted back to the user to ensure the integrity of the agent’s instructions and the security of the overall environment. This evolution marks a departure from purely decentralized interactions, as users must now trust the underlying code and the decision-making logic of the AI agents they deploy within the financial ecosystem.
Outlook
The data from 2025 suggests that the integration of automation and AI within the cryptocurrency ecosystem is becoming a standard feature of the trading landscape. The 97% year-over-year growth in grid trading volume at HTX, combined with the doubling of capital allocated to these strategies, points to a market that is increasingly comfortable with algorithmic execution. The particularly high growth in stablecoin grid trading at 352% indicates that users are finding high utility for these bots in lower-volatility environments, especially when compared to the 122% growth for major cryptocurrencies.
As Coinbase continues to refine its AI agent offerings, such as the Payments MCP introduced in October 2025 and the Based Agent tool from the previous year, the distinction between manual trading and autonomous onchain activity is expected to blur. The ability for AI agents to conduct swaps, staking, and payments without the use of API keys represents a significant technical milestone for the industry. While early adoption appears strong, with over a third of surveyed users open to AI-managed holdings, the industry continues to weigh the benefits of efficiency against the security risks highlighted by experts. The continued dominance of major exchanges like HTX—which remains a top-10 platform for volume and liquidity—ensures that these automated tools will have a broad user base. Moving forward, the focus is likely to remain on how these AI systems can balance autonomous decision-making with the security requirements of the blockchain environment.
About the Author
