Home / Business and Economy / AI Agents Get a Safety Net: Control Before Action
AI Agents Get a Safety Net: Control Before Action
19 Oct
Summary
- AI actions are checked before execution for policy compliance.
- Financial institutions can approve or block AI actions with signed records.
- Cortx Sentinel is available at no extra cost to Xnode Cortx customers.
Xnode has launched Cortx Sentinel, an enhancement to its Xnode Cortx AI Control Plane, designed to vet AI agent actions prior to execution. This new capability empowers agents that adhere to firm policies with greater autonomy, while flagging and restricting those that do not. Financial institutions can now effectively govern AI-driven operations, including payments, by approving or blocking agent actions in real-time.
Cortx Sentinel ensures that each AI action is verified against established policies. For instance, a payment exceeding an agent's limit would be held for secondary approval, with all decisions meticulously logged for auditor verification. Security teams can also halt rogue agents within seconds. This system operates within the customer's network, addressing critical risk management concerns for banks.
Available on AWS, Microsoft Azure, and Google Cloud marketplaces, Cortx Sentinel is provided at no additional charge to existing Xnode Cortx customers. The company is offering a 30-day free trial in customers' own environments. Xnode Cortx has seen significant growth, with licensed users increasing by 50% and monthly token volume rising into the billions.
