What are the differences between the four risk levels illegal, fraud, risk, and warning when using the Web component to access a page? Does the risk levels can be customized?

THREAT_ILLEGAL and THREAT_FRAUD indicate that the access is forbidden and the Continue button is unavailable.
THREAT_RISK indicates that the access is forbidden, but the Continue button is available.
THREAT_WARNING indicates that the Web kernel does not intercept the request, displays a warning message, and there is no Continue button.

Currently, the risk-based access control cannot be customized, and no related callback is provided.

Reference

ThreatType