Personally Identifiable Information
Names, email addresses, SSNs, ID numbers, phone numbers
Example: Employee pastes a customer CSV into ChatGPT for analysis, Aona blocks and notifies.
Enterprise AI Firewall
AI Firewall forEnterprise
Protect sensitive data before it reaches AI tools. Aona applies real-time blocking or redaction to supported AI interactions through a deployed browser plugin or native app.
POLICY INPUTCustomer record detectedPolicy checkSubmission paused
POLICY OUTCOMESubmission paused
Synthetic policy response on a supported evaluation path.
Synthetic example. Validate the client and policy path in scope.
48%
of Cisco survey respondents entered non-public company information into GenAI (2024)
10,000+
AI tools in the catalogue
IT
managed endpoint deployment
2
endpoint options: browser plugin or native app
Cisco Data Privacy Benchmark Study 2024, Figure 18
How Aona AI Firewall Works
Discover AI use, apply data guardrails and enforce policies on covered endpoints.
- 01
Detect
Discover covered AI use
Discover approved and unapproved AI use observed on covered endpoints. Review which tools employees use and the usage and policy events captured by your deployment.
10,000+ AI tools in the catalogue - 02
Block
Real-time DLP rules
On supported input paths, Aona inspects prompts before submission. Configure rules for PII, financial data, source code and custom data types to block or redact sensitive content.
Inspect sensitive data before submission - 03
Enforce
Policy by team and role
Apply granular AI acceptable use policies by team, department, or individual role. Engineers may use code-aware AI tools. Finance teams may be restricted from AI access to financial models. One platform, unlimited policy flexibility.
Role-based policies on covered endpoints
What Aona AI Firewall Blocks
Real-time protection for every sensitive data type, with employee coaching at the moment of risk.
Financial Data
Account numbers, payment records, revenue figures, forecasts
Example: CFO's assistant drafts board report using Copilot with unreleased earnings, Aona redacts.
Source Code
Proprietary code, API keys, infrastructure configs, trade secrets
Example: Developer pastes internal auth module into GitHub Copilot, Aona blocks and coaches.
Customer Data
CRM records, support tickets, deal information, contracts
Example: Sales rep asks Gemini to summarise a deal with full customer detail, Aona redacts contact data.
Confidential Documents
Board minutes, M&A docs, HR records, legal correspondence
Example: HR manager uploads performance review to Claude for summarisation, Aona intercepts.
Aona AI Firewall vs Legacy DLP
Traditional DLP tools were built before generative AI existed. Aona was built for it.
| Capability | Aona AI Firewall | Legacy DLP |
|---|---|---|
| Real-time prompt inspection | ✓ | × |
| AI-native classification | ✓ | × |
| Works with SaaS AI tools | ✓ | × |
| Employee coaching at point of risk | ✓ | × |
| Role-based AI policies | ✓ | × |
| Shadow AI discovery | ✓ | × |
FAQ
Frequently Asked Questions
What is an AI firewall?
An AI firewall is a security layer for employee AI use. It inspects prompts and supported file uploads for sensitive information such as PII, financial records and source code, then applies configured blocking or redaction before submission. Aona uses a deployed browser plugin or native app, with coverage depending on the supported AI tool, client and input path.
How does an AI firewall differ from a traditional DLP?
Traditional DLP (Data Loss Prevention) was designed for email, file transfers, and cloud storage, not AI tools. It operates reactively, often flagging violations after data has already been transmitted. An AI firewall is purpose-built for generative AI: it inspects prompt content in real time, understands AI-specific data flows, coaches employees at the moment of risk, and enforces policies across SaaS-based AI tools that traditional DLP cannot reach.
Does Aona AI Firewall work with ChatGPT and Microsoft Copilot?
Aona supports AI interactions with services including ChatGPT, Microsoft Copilot, Google Gemini and Claude, with controls varying by client and input path. Its catalogue covers 10,000+ AI tools for discovery and risk context, not uniform enforcement across every tool. Your IT team deploys the browser plugin or native app and confirms the supported coverage for your rollout.
Can I set different AI firewall rules for different teams?
Yes. Aona AI Firewall supports granular policy controls by team, role, department, and tool. For example, your engineering team can be permitted to use AI coding assistants with source code while your sales team's use of ChatGPT is restricted from sharing customer data. Policies can be configured in the Aona dashboard and applied instantly across your organisation.
Get started
Stop Data Leaks Before They Happen
See how Aona protects supported AI interactions. Plan browser-plugin or native-app deployment with your IT team and confirm coverage for the tools your employees use.