See and stop AI data leaks across every tool
Let's talk
per seat · no minimums
Give your team a safe AI they can actually use
Let's talk
per seat · no minimums
Govern all your AI — including AI agents
Let's talk
per seat
AI data-loss prevention, shadow-AI discovery, governed AI chat and agentic protection — by plan.
| StarterProtect | Most PopularProProvide | EnterpriseGovern | |
|---|---|---|---|
Sensitive‑data detection (50+ types) | |||
| Secrets & API keysAWS, Azure, GCP, private keys, JWTs, GitHub, Stripe, DB connection strings… | |||
| Government & national IDsUS SSN/ITIN/EIN, UK, Canada, Australia, Germany, France, Spain, Italy, Netherlands | |||
| Identity documents (passport, driver's license) | |||
| Payment & financialCards (Luhn‑validated), IBAN, bank/routing, crypto wallets | |||
| Healthcare / PHIMedical record & insurance IDs, DOB, NPI, DEA | |||
| Classified & controlled markings (CUI, NOFORN…) | |||
| Internal network data & secrets in code | |||
| Contact PII (email, phone) | |||
| AI/ML detection — names & free‑text (NER) | |||
| Custom sensitive terms (your own data) | |||
Prevention | |||
| Block · Warn · Coach in real time | |||
| Per‑category policy (block / warn / off) | |||
| Reversible redaction — the AI sees only masked data | |||
| Response inspection — scans the AI's replies too | |||
Governed AI chat | |||
| Inspected AI chat (block never reaches the AI) | |||
| AI model | Efficient | Advanced | Frontier |
| Choice of provider (ChatGPT & Claude) | |||
| Streaming replies + chat history | |||
| Live web search | |||
| Upload & analyze documents (PDF, DOCX, CSV…) | |||
Shadow‑AI discovery | |||
| Browser sightings + dashboard (per‑vendor, top users) | |||
| DNS / proxy‑log import — catches desktop & CLI tools | |||
| Provider API‑key audit | |||
| Shadow‑AI trend history | |||
Reporting & analytics | |||
| Security overview — leaks prevented, detection rate | |||
| Event log (filterable, masked previews) | |||
| AI‑usage report (by provider & model) | |||
| DLP‑prevention report (blocked / warned / redacted trends) | |||
| Timeline, top categories, users‑by‑risk, peak hours | |||
| Policy‑effectiveness report | |||
| CSV export for SOC 2 / audit | |||
Compliance & audit | |||
| Audit log (logins, policy/config changes, IP) | |||
| Warn‑override audit trail + masked‑only storage | |||
| Regulation presets — HIPAA, PCI, GDPR, GLBA, DoD/CUI | |||
| Incident email alerts (block / override) | |||
| Configurable data retention | |||
Admin, access & identity | |||
| SSO (Microsoft Entra) + magic‑link login | |||
| Multi‑tenant isolation + role‑based access | |||
| User management + team invitations | |||
| Cross‑browser extension (Chrome, Edge, Firefox) | |||
| Zero‑touch mass deployment (GPO / Intune force‑install) | |||
Agentic / MCP protection | |||
| MCP gateway for Claude Cowork & Code | |||
| Block / monitor / redact modes + connector management | |||
For developers | |||
| Inspection API — embed DLP checks in your own apps | |||
Support | |||
| Support | Standard | Standard | Priority |
Sold through TD Synnex and our MSP partners, or direct — per seat, no minimums. Book a call and we'll tailor it to your environment.
A single user running heavy queries on Claude Opus can cost more in a day than an entire team’s monthly subscription. Caps protect predictable pricing — yours and ours. If your users regularly hit a cap, that’s the signal to upgrade.
Also, we don’t sell unlimited at any tier. We’d have to charge several hundred dollars per user per month to underwrite the worst case, and most teams wouldn’t come close to using it. Enterprise with volume pricing is the right answer for very heavy use. We can customize usage tiers for your organization upon request.
Yes. Enterprise customers can connect approved OpenAI, Anthropic, Google, or other provider API keys while still using AILeakShield’s inspection engine, policy controls, audit logging, and custom data protections.
Blocked content never leaves AILeakShield — caught by the inspection engine in memory and never sent to any AI provider. We never store sensitive data and we don’t allow it to get to the AI providers either. Allowed content is forwarded to OpenAI, Anthropic, or Google through your selected model. We retain only a redacted audit record (the decision, the category, a masked preview) — never the raw prompt content. Blocked-tier events store only a category placeholder (e.g., ), not the original text.
Yes. Customers commonly start with secure AI portal access and later expand into browser-wide protection, shadow AI monitoring, advanced models, and enterprise policy controls as adoption grows.
No. Users login to app.aileakshield.com like they would any other AI solution. They select their preferred AI model and use it like any other chat.
There is an optional browser monitoring solution included with our pro and enterprise tier. This allows you to monitor for shadow AI throughout your environment. It also allows you to apply policy to the browser that you have in the secure workspace. Your IT team pushes the extension through Microsoft Intune, Chrome Enterprise, Active Directory Group Policy, or JAMF — exactly the same way they push any other managed extension. Users see the AILeakShield icon appear in their toolbar, sign in once via your existing Microsoft SSO (typically silent), and then forget it exists until it blocks a prompt.
The five built-in presets (HIPAA, PCI, GDPR, GLBA, Defense / CUI) cover most enterprise needs out of the box. The underlying policy engine is fully tunable per category — dial detection up or down to match internal frameworks (ISO 27001, NIST 800-171, FERPA, SOX, etc.). Enterprise customers can also add custom regex patterns for organization-specific data (project codenames, customer identifiers, internal product names). If you require a framework not already covered, just let us know, we add it as a feature request and respond quickly.
Three things you can’t get on any other tier: (1) access to the most advanced AI models (GPT-5, Claude Opus 4.7), (2) the enhanced shadow-AI monitoring suite — live dashboard, DNS log import, API-key audit, (3) custom regex patterns for organization-specific data — plus longer audit retention, higher per-user message caps, signed BAA for HIPAA workloads, volume pricing, and a named customer success manager.
AILeakShield helps your team use the AI tools they already want with prompt protection, Microsoft SSO, usage controls, daily prompt text deletion, and enterprise-ready security controls.
