Xage don expand Zero Trust for AI make agent access dey jailbreak-proof across cloud, SaaS, and edge

Xage Security talk say companies need stronger Zero Trust for AI as AI agents dey move from experiments go production. Di company don announce say dem expand “Zero Trust for AI” controls across cloud, SaaS, on-prem and edge, aim na make am “jailbreak-proof” security. Key update: Xage dey focus on deterministic visibility and enforce boundaries on wetin AI agents fit do—on system level, no be only prompt monitoring. Di platform add two core components: Agent Sentry (e wrap agents and dey monitor inputs/outputs) and Resource Gateway (e sit for front of critical resources to govern agent interactions). Together, dem wan block unauthorized actions in real time and produce detailed audit logs across full AI interaction chain. Xage highlight agent-enabled risks like prompt injection manipulation, unintended actions, and possibility of sensitive-data exfiltration—especially when agents get wide API and operational access. E still add anomaly detection (e.g., unusual activity spikes or unauthorized writes) and lifecycle management using unique agent identities for role-based policies and faster termination of compromised agents. For crypto traders, na cybersecurity/enterprise software update dis no be direct token catalyst. But e fit support sentiment around “AI infrastructure” security as institutions ramp up agent deployments, though e unlikely to move big coins on its own.