Trend Micro Vision One AI Application Security

Trend Micro Vision One AI Application Security’s AI Guard feature uses a configurable policy to identify risks in AI Applications, such as:

Prompt injection attacks
Toxicity, violent, and other harmful content
Sensitive Data

Setup

Create a new Vision One API Key with permissions to Call Detection API
See the AI Guard Integration Guide for details around creating your policy

Colang v1:

1 # config.yml
2 
3 rails:
4   config:
5     trend_micro:
6       v1_url: "https://api.xdr.trendmicro.com/v3.0/aiSecurity/applyGuardrails"  # Trend Micro AI Guard Endpoint
7       api_key_env_var: "V1_API_KEY"
8       application_name: "my-ai-app"  # Required: Application identifier (max 64 chars, alphanumeric, hyphens, underscores)
9       # Optional:
10       detailed_response: true  # Set to true for detailed AI Guard results
11       # For other regions, use: https://api.{region}.xdr.trendmicro.com/v3.0/aiSecurity/applyGuardrails
12       # where region is: eu, jp, au, in, sg, or mea
13   input:
14     flows:
15       - trend ai guard input
16 
17   output:
18     flows:
19       - trend ai guard output

Colang v2:

1 # config.yml
2 colang_version: "2.x"
3 rails:
4   config:
5     trend_micro:
6       v1_url: "https://api.xdr.trendmicro.com/v3.0/aiSecurity/applyGuardrails"  # Trend Micro AI Guard Endpoint
7       api_key_env_var: "V1_API_KEY"
8       application_name: "my-ai-app"  # Required: Application identifier (max 64 chars, alphanumeric, hyphens, underscores)
9       # Optional:
10       detailed_response: true  # Set to true for detailed AI results
11       # For other regions, use: https://api.{region}.xdr.trendmicro.com/v3.0/aiSecurity/applyGuardrails
12       # where region is: eu, jp, au, in, sg, or mea

# rails.co
import guardrails
import nemoguardrails.library.trend_micro
flow input rails $input_text
    trend ai guard $input_text
flow output rails $output_text
    trend ai guard $output_text