Riscosity Data Flow Governance 作者： Riscosity

Riscosity is an enterprise data security platform that gives security and compliance teams visibility and control over how employees use AI tools. As AI assistants become standard in the workplace, sensitive data — customer PII, financial records, source code, credentials — is routinely typed into chat prompts and sent to third-party AI providers. Riscosity stops that.

What the extension does:

Real-time DLP — detects sensitive data (PII, credit cards, SSNs, API keys, health information) as employees type into AI chat boxes, and highlights or anonymizes it before the prompt is submitted
Prompt logging — captures and logs every AI prompt to the Riscosity compliance dashboard, giving Privacy Officers and CISOs a full audit trail
File upload scanning — intercepts files attached to AI chats and scans them for sensitive content before upload is permitted
Injection detection — flags prompts that show signs of prompt injection attacks using on-device WASM-based heuristics
Site blocking — enforces organisation policy by blocking access to unauthorised AI services or websites entirely
Account awareness — distinguishes between personal and corporate AI accounts so different policies can be applied to each

For IT admins: The extension supports zero-touch deployment via Firefox enterprise policy (policies.json), allowing API keys and configuration to be pre-seeded across a managed fleet without end-user action.

Requirements: An active Riscosity organisation account and API key. Visit riscosity.com to request a demo or trial.