Skip to main content
The Verdict System defines how Scanurl.ai classifies a URL after analysis and helps applications decide how to respond based on risk levels.

​
Overview

Every scan returns a final_verdict along with a confidence level and a risk_score. The verdict represents the overall risk classification based on multiple layers of analysis, including URL patterns, page behavior, visual signals, and infrastructure data.

​
How to Interpret Each Verdict

​
Trusted

The URL appears safe based on all available signals. Recommended Usage:
  • Allow access normally
  • No warning required

​
Low Risk

The system detected minor signals, but nothing strongly suspicious. Recommended Usage:
  • Allow access normally
  • Optional logging or monitoring

​
Moderate Risk

Some indicators suggest potential risk. Recommended Usage:
  • Show a warning before user proceeds
  • Log for review

​
Elevated Risk

Strong indicators of suspicious or deceptive behavior. Recommended Usage:
  • Warn users clearly
  • Consider restricting access

​
Potential Fraud Risk

High probability of malicious intent. Recommended Usage:
  • Block access
  • Prevent user interaction
  • Log as high-risk event

​
Inaccessible

This verdict is returned when the system is unable to complete the analysis for the given URL. This typically occurs when the target website is not reachable at the time of scanning. Possible reason:
  • The target server is down or not responding
Important Note: The Inaccessible verdict does not indicate that the URL is safe or unsafe.
It only means that the system could not analyze the page at that moment. Recommended Usage:
  • Retry the scan after some time
  • Treat the result as unknown
  • Avoid allowing critical actions based on this verdict alone

​
Known Threat

If the URL or domain matches an entry in the internal threat database, the system may immediately classify it as a known threat. In such cases:
  • The scan may be short-circuited
  • Deep analysis may be skipped
  • Verdict is based on existing intelligence
Recommended Usage:
  • Block immediately
  • Do not allow user interaction
  • Treat as confirmed malicious
For most applications:
  • Allow → Trusted / Low Risk
  • Warn → Moderate Risk
  • Restrict → Elevated Risk
  • Block → Potential Fraud Risk / Known Threat