Your company protects a public web application with an Azure Front Door Premium WAF policy running in Prevention mode with the Microsoft Default Rule Set (DRS) enabled. After deployment, users report that a legitimate form submission is blocked. Reviewing the WAF logs, you find the requests are matched by a specific SQL injection managed rule because a text field named 'comments' contains the string 'or 1=1' as part of normal user input. You must stop blocking these legitimate submissions while keeping the WAF in Prevention mode and maintaining protection against genuine attacks on all other fields and rules. What should you do?