We build safety into the foundation of our platform, ensuring that AI agents operate within clear boundaries while enabling powerful new commerce experiences.
Core tenets that guide how we build and operate our platform.
Every agent action can be reviewed and approved by humans. Users maintain full control over what their agents can do.
Configurable transaction limits ensure agents operate within user-defined boundaries for all financial activities.
Advanced ML models detect and prevent suspicious transactions before they occur, protecting users and merchants.
Complete transaction history and decision logs provide full transparency into agent behavior.
Our platform includes multiple layers of protection to ensure safe operation of AI agents in financial contexts.
Read our research papers and documentation on safety practices.
Read Safety Research