Ripple Expands GSmart AI Agents Across Its Corporate Treasury Platform to Flag Risks and Recommend Actions
Ripple has spent the past year building out a corporate treasury business almost from scratch, and on September 10 it took the next step by loading that platform with AI agents built to watch a company’s cash positions and flag problems before a human ever has to go looking for them.
What actually launched
The expansion adds new capabilities to GSmart, the AI toolset embedded across Ripple Treasury, the platform Ripple built on top of GTreasury following its roughly $1 billion acquisition of the treasury-management firm in October 2025. The new agents work across five core functions: liquidity forecasting, cash management, risk assessment, real-time reconciliation, and financial reporting. Rather than simply generating dashboards, the agents are designed to actively surface anomalies, draft recommended actions, and explain the reasoning behind them in plain language.
Why Ripple built in a governance layer
Handing software the ability to recommend financial moves invites an obvious worry: what stops it from doing something a finance team never approved? Ripple’s answer is a component called Knowledge Studio, which acts as the rules engine sitting underneath GSmart. Treasury teams use it to define their own organizational policies and controls, and every action the AI proposes is checked against those rules before it can be escalated to a human for sign-off. Ripple has been explicit that no automated action executes without manual authorization — the agents recommend, they do not execute.
A conversational layer on top
The rollout also includes Analytics Studio, home to a feature called Ask GSmart. It gives treasury staff a chat-style interface for querying their own financial data directly, pulling answers and insights without needing to build a custom report first. Ripple describes the overall approach as “treasury-native AI” — tools designed around the specific controls and workflows corporate finance departments already use, rather than a generic AI layer bolted on afterward.
Early adoption numbers
Ripple disclosed some early usage figures alongside the announcement: roughly 60% of eligible customers have already switched on Risk Insights, the feature that hunts for exposure anomalies and policy breaches, while 44% of qualified customers are using Forecast Insights to compare projected cash flows against what actually materializes. Those numbers suggest the underlying GTreasury customer base — built over decades before Ripple’s involvement — has been reasonably quick to try the new AI layer once it became available.
The bigger picture
Ripple’s entry into corporate treasury is still relatively new. The GTreasury acquisition gave it access to an enterprise customer base far removed from crypto-native trading desks, and in April 2026 Ripple layered in Digital Asset Accounts and Unified Treasury, letting corporate customers manage fiat and digital liquidity — including XRP and Ripple’s RLUSD stablecoin — from the same dashboard. The GSmart AI expansion is the next layer on that stack, and it puts Ripple in more direct competition with legacy treasury-software vendors that are racing to add their own AI features for the same corporate finance audience.
Note: the second embedded post above is Ripple’s April 2026 announcement of Digital Asset Accounts and Unified Treasury, included here as directly relevant background context rather than as coverage of the September 10 GSmart expansion itself.
Enterprise AI tools and treasury platforms carry their own operational and counterparty risks, and nothing here constitutes financial or investment advice. Always verify claims independently before making decisions involving corporate funds or crypto assets.
Want to understand how platforms like this fit into the bigger picture of digital-asset infrastructure? Start with coin680’s Bitcoin Academy.