I am almost certain they violated their own policies for telling you which word triggered their filters.

Regular KYC procedure usually involves using a blacklist of disallowed words, and then if anyone triggers the filter, you block them, and ask them to submit any and all documentation they have for all recent transactions (but importantly, you dont tell them the transaction or word that triggered the check). Someone then reviews the documentation, and unless it explains the blacklisted word, the account ban stands.