Crypto exchange Bitget said last Wednesday that it has launched a dedicated account structure for its AI agent, GetClaw, allowing the system to place and manage live trades autonomously inside an isolated sub-account.
The move extends Bitget’s existing Agent Hub setup by letting GetClaw act within a ring-fenced environment rather than simply generating trade suggestions or analysis. Users can define strategies in natural language, and the agent then monitors markets continuously and manages positions in real time without requiring manual approval at each step, according to the company.
The product closes the loop from user instruction to analysis to execution to ongoing position management.
The dedicated sub-account separates agent-driven activity from user-controlled assets, a design Bitget frames as adding clearer boundaries and permissions for autonomous execution.
“Sooner or later emerging financial markets are going to be filled with AI agents trading on behalf of users. We’re preparing the infrastructure to run this on scale,” said Gracy Chen, CEO at Bitget.
Rollout sequence
The launch is the third step in a product roadmap that began in February. Bitget launched Agent Hub on February 13, establishing standardized infrastructure for AI agents to access market data and execution tools. In March, the company expanded Agent Hub with five analytical AI Skills and 19 integrated data tools, connecting analysis and execution inside one environment.
Last Wednesday’s update adds autonomous execution inside a dedicated account, completing a progression the company describes as moving from access, to intelligence, to independent execution.
Bitget positions the broader direction as building an ‘agent-native exchange,’ where AI systems operate as first-class participants alongside human traders inside the same infrastructure.
