Ripple Expands $1B Treasury Bet With AI-Powered Agents
Ripple is expanding its corporate treasury platform with AI agents that can track cash, monitor risk, review forecasts and recommend financial decisions. Despite the increased automation, employees must approve every action before it is executed.
The new GSmart features are integrated into Ripple Treasury, formerly GTreasury, which Ripple acquired in October as part of its $1 billion push into the corporate treasury market.
The AI tools can monitor treasury activities such as liquidity management, forecasting and reconciliation. They can flag potential issues, suggest a course of action and identify the company policy supporting their recommendations. Ripple said no action is automatically carried out without human approval.
The company is keeping financial calculations separate from AI functions. Deterministic software performs the calculations, while AI is used to interpret policies, identify patterns and explain the reasoning behind recommendations.
GSmart was first introduced in June 2025, before Ripple acquired GTreasury. The latest update extends the technology across more treasury operations and adds Knowledge Studio, a policy-focused layer designed to provide additional context for the AI.
Ripple has also introduced Ask GSmart, allowing finance teams to query corporate financial data through the platform.
The company said adoption of its existing tools remains strong. About 60% of eligible customers have enabled Risk Insights, while 44% are using Forecast Insights.
The rollout further advances Ripple’s strategy following its GTreasury acquisition. Earlier this year, CoinDesk reported that Ripple had started integrating digital assets into its treasury software, allowing companies to manage digital holdings alongside conventional cash.
Ripple Treasury can now manage both traditional and digital assets, combining conventional corporate treasury functions with AI-powered analysis and blockchain-based financial capabilities.
Share this content:













