Enforce ethical AI usage with automated protections for content moderation, privacy, and compliance in your Lyzr agents.
0.4
โจ Agents will automatically block or filter responses that exceed the toxicity threshold.
0.3
๐ Especially useful in agents interacting with untrusted or anonymous users.
๐ง Useful for domain-specific AI assistants with strict focus.
๐ก Blocked keywords will be filtered out or replaced during processing.
Data Type | Description | Options |
---|---|---|
Credit Card Numbers | Detects 13โ16 digit card numbers | Disabled / Blocked / Redacted |
Email Addresses | e.g., john@example.com | Disabled / Blocked / Redacted |
Phone Numbers | International and local formats | Disabled / Blocked / Redacted |
Names (Person) | Common personal name patterns | Disabled / Blocked / Redacted |
Locations | City, state, country, address mentions | Disabled / Blocked / Redacted |
IP Addresses | IPv4 / IPv6 addresses | Disabled / Blocked / Redacted |
Social Security Numbers | U.S. SSN format: XXX-XX-XXXX | Disabled / Blocked / Redacted |
URLs | Any web address patterns | Disabled / Blocked / Redacted |
Dates & Times | Recognizable temporal references | Disabled / Blocked / Redacted |
๐ These controls help you comply with GDPR, HIPAA, and other data protection standards.
Use Case | Responsible AI Features Used |
---|---|
Customer Support Chatbot | Toxicity Filter, Secrets Masking, PII Redaction |
HR Agent for Internal Use | Allowed Topics, Blocked Keywords, PII Redaction |
Public-Facing Financial Assistant | Prompt Injection Detection, Banned Topics, URL Redaction |
Legal Document QA Bot | Secrets Filter, Credit Card Blocking, Topic Control |
โ๏ธ Changes take effect immediately for all new interactions.