Auto-suggestions can sometimes be harmful, as they may include offensive or violent content. Sometimes, services that I have already blocked still appear in the suggestions. I want the ability to filter out certain suggestions. To address this, I propose a censorship feature using regex in the auto-suggestion system. When I provide my token to the auto-suggestion API, it should return suggestions based on a predefined regex list, ensuring that unwanted content is excluded.
Just like Google's auto-suggestion reporting system, users should be able to remove unwanted suggestions from the list. Furthermore, an AI-powered regex generation feature could be considered. For users unfamiliar with regex, the AI could automatically convert specific suggestion patterns into regex rules. Additionally, similar to the existing "Raise" feature, users could choose regex patterns based on aggregated statistics from the entire user base.
