ShadowLock is a shadow AI detection and governance platform that gives MSPs and IT teams real-time visibility and control over how employees use AI tools, before sensitive data leaves the endpoint.
What is ShadowLock?
ShadowLock detects and blocks unauthorized AI use on Windows endpoints by combining a silently deployed agent, a browser extension, and a multi-tenant dashboard. It monitors AI website visits, browser extensions, desktop AI apps (including local LLMs like Ollama and LM Studio), and Microsoft 365 OAuth-connected AI apps via Graph API. The platform logs all AI activity and enforces allow/warn/block policies without capturing sensitive data content — only metadata and data types are recorded.
Key Features
- AI Website Detection & Blocking — Automatically identifies known AI domains (ChatGPT, Claude, Gemini, etc.) and applies block/warn/allow policies before any data is pasted, with an automatically updated domain list.
- Sensitive Data Interception — Stops paste events and file uploads containing PII, credentials, SSNs, or card data before content reaches the AI tool; classification happens entirely within the browser.
- Personal Account Detection — Reads the signed-in identity on AI sites and blocks prompts from personal or unauthorized accounts until a corporate account is verified, without per-site setup.
- Browser AI Lockdown — Disables the native AI features built into Chrome (Gemini), Edge (Copilot), Brave (Leo), and Firefox, with drift detection if re-enabled.
- Google Search AI Mode Control — Blocks Google Search's conversational AI Mode while allowing standard search to proceed.
- Desktop AI App Detection — Surfaces usage of offline AI tools, local LLMs, and developer apps like Ollama and LM Studio that run outside browser controls.
- Microsoft 365 OAuth AI App Detection — Connects to each customer's M365 tenant via Graph and scans for AI apps granted OAuth access; new connections trigger critical alerts automatically, no endpoint required.
- Multi-Organization MSP Dashboard — Provides cross-org risk view, alert workflows, device inventory, and policy management with exportable reports for clients and auditors.
Who is it for?
- MSPs and partners — Govern AI risk across all customers from one dashboard, push policy changes to hundreds of endpoints via any RMM, and generate customer-ready audit reports.
- IT admins — See which AI tools employees use, what data they paste, and which extensions are risky; configure allow/warn/block without complex rule engines and get alerts that matter.
- Compliance and security teams — Answer auditor questions (e.g., "How do you prevent PII from entering public AI systems?") with documented enforcement evidence and exportable logs.
What can you do with ShadowLock?
- Block personal-account ChatGPT usage: Detect when employees use ChatGPT with a personal Google account instead of corporate SSO, and block the prompt until a managed account is verified.
- Stop sensitive data from reaching AI tools: Intercept paste events containing SSNs or credit card numbers before they reach Claude or Perplexity, with the data type logged locally and never transmitted.
- Enforce AI data-sharing opt-out: Read each AI tool's real "train on my data" setting and block prompts until it's switched off for ChatGPT, Claude, Perplexity, Le Chat, Copilot, and Grok.
- Audit AI browser extensions: Scan all Chrome and Edge extensions per endpoint, flag known AI sidebars and tools, and highlight unknown extensions with high-risk permissions.
