Service A acts as a trusted subsystem for a shared database. The database contains sensitive information and performs strict validation on all incoming data modification requests. In case of any invalid input values, the database throws detailed error messages that are required for debugging purposes and are automatically relayed back to service consumers by Service A. Recently, while going through the access logs of the database, it has been reported that attempts have been made to connect to the database from outside the organization. What can be done to prevent such attacks while preserving the existing database debugging requirements?