Binance, the world’s largest cryptocurrency exchange, launched Agent OS, a platform that enables AI agents to analyze financial markets and execute trades on users’ behalf. The platform integrates Binance’s existing tools, including APIs, Wallet Agentic Hub, and Model Context Protocol (MCP), while also supporting external AI applications such as ChatGPT and Claude Code. Users can authorize agents to access market data and perform trades, but they retain control over permissions and limits. Binance’s vice president of product, Jeff Li, emphasized that users must determine what agents can access and do, with granular control over their activities.
Binance’s approach involves sub-accounts, which users can assign to agents and configure for specific trading activities like spot or futures trading. Withdrawals from these sub-accounts are blocked by default, creating a sandbox environment for agent activity. Users can also set whether agents require approval for each trade or can execute orders autonomously once permissions are configured. Binance does not impose a separate cap on trading limits, meaning the amount transferred into the sub-account effectively determines the limit.
Binance’s Agent OS allows users to monitor trading activity but has limited visibility into the reasoning behind an agent’s decisions, as the AI reasoning occurs outside Binance’s systems. The company’s existing security and anti-money-laundering policies for subaccount APIs apply to Agent OS. Trading is the first use case, but the platform is also designed to connect agents to payments and on-chain activities, with daily transaction limits set by Binance for different types of transactions.
Source: techcrunch
