Binance, a major cryptocurrency exchange, has launched Agent OS, a platform designed to allow AI agents to analyse markets and execute trades on behalf of its more than 300 million registered users. This development brings autonomous AI directly into the management of real money.
Agent OS connects AI applications and agents to Binance's financial infrastructure, including existing tools like Binance APIs and Binance Wallet Agentic Hub, alongside new support for its Model Context Protocol (MCP). The platform is compatible with various AI tools, such as OpenAI's ChatGPT and Codex, Anthropic's Claude Code, and Cursor.
Users can authorise agents to access market data, view account information, and execute trades. Binance places the responsibility on users to manage these agents, allowing them to decide what agents can access and trade, and to set limits on their activities. This control is primarily managed through dedicated sub-accounts, which users can configure for specific activities like spot or futures trading. Withdrawals from these sub-accounts are blocked by default.
Users can also choose whether an AI agent requires approval for each order or can execute trades autonomously once permissions are set. Binance does not impose a separate cap on how much an AI agent can trade or lose within a sub-account, meaning the amount transferred into the sub-account acts as the limit. For Agentic Wallet transactions, Binance has set daily limits, such as $50,000 for regular swaps and $100,000 for DeFi transactions.
Jeff Li, Vice President of Product at Binance, stated that the reasoning behind an agent's trade decisions occurs outside Binance's systems. Binance can monitor the resulting trading activity but has limited visibility into the decision-making process itself. Other crypto exchanges, including Kraken, Coinbase, and OKX, have also introduced similar functionalities for AI agents.