ZeroDrift launches service to check agent-generated messages against company policies
ZeroDrift Inc., a startup that automates compliance for artificial intelligence communications, today introduced Guard for Agents, a service that lets developers turn written company policies into enforceable rules to check AI agents’ communications before they reach customers. The launch extends the company’s compliance technology into agent workflows through an application programming interface…
ZeroDrift Inc., an artificial intelligence compliance startup, has launched a new service called Guard for Agents. The service enables developers to convert company policies into enforceable rules that can check AI agents' communications before they reach customers. Guard integrates with various platforms like Notion Labs, Linear Orbit, and Google Drive to import policy documents.
Developers can embed these checks into applications through an API, checking outgoing responses for violations and providing details about any infractions. Customers can decide on the handling of violations, ranging from blocking messages to human review. ZeroDrift's small language model identifies rule sets from policy documents, enabling developers to configure and enable or disable rules as needed.
The service is intended to operate at the pre-send layer, intercepting messages before they are sent. ZeroDrift's technology can also handle communications from conventional CRM and marketing systems. The company emphasizes the speed and cost-effectiveness of their solution compared to using general-purpose models for review.
Written by urgent.news from SiliconANGLE's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
