What happened
On Nov 9, 2026 Varonis announced three new Atlas capabilities, available now: Browser Protection (browser extension that finds personal/unsanctioned AI usage, blocks unsanctioned AI sites, and applies data-aware guardrails to prompts/responses/attachments), Endpoint Protection (discovers AI apps, coding agents like Claude Code, local model files, and MCP servers on laptops/desktops; integrates with CrowdStrike, SentinelOne and Microsoft Defender), and an AI Gateway that fronts LLM and MCP traffic with model routing, token-spend caps and prompt/response/tool guardrails. It also expanded Atlas Agent Intent with a new Agent Manifest defining each agent's purpose, driving intent-based access control (IBAC).
Why it matters
Varonis connects AI activity directly to its Data Security Platform, letting a data-security vendor put consistent policies on AI usage across browser, endpoint, cloud and apps — closing the gap between 'I can see AI' and 'I can control data exposure.' Available now from a significant NASDAQ-listed vendor gives it broad adoption reach.
Applicability
Enterprises with Varonis DSP or those seeking to rein in shadow AI and control data access from AI use should evaluate these Atlas capabilities; useful for teams needing browser/endpoint AI discovery plus gateway-level control in one platform.