Merge Agent Handler adds LLM-based DLP guardrails
Agent Handler has introduced LLM-based Data Loss Prevention (DLP) guardrails designed to inspect both inputs and outputs of tool calls for sensitive information. Security teams can toggle this protection layer on or off based on their specific governance requirements, ensuring agentic interactions with enterprise tools remain compliant and secure.
Adding real-time LLM-based DLP to AI agent tool calls is a crucial evolution for enterprise deployment, where sensitive data leakage remains a top barrier to adoption.
- –Inspects both incoming tool inputs and outgoing tool responses for sensitive data exposure.
- –Gives security and IT administrators configurable toggle controls to meet custom compliance requirements.
- –Mitigates risks associated with autonomous tool invocation and unauthorized data exfiltration across enterprise apps.
DISCOVERED
46d ago
2026-08-05
PUBLISHED
46d ago
2026-08-05
RELEVANCE
AUTHOR
merge_api