When someone complains about the use of a trademarked keyword G places the word or phrase in a database.
Apparently they periodically scan all ad text for hits on the trademark. G attempts to figure out the context of the ad and if context is related to trademark then violation is flagged. Seems their algo to figure out context needs some more brains.
My ad has to do with database design and programming techniques - not horse feed! :)
Seems a bit paranoid to me, but we just modify the ad to something similar. A number of times, the new ad even outperformed the original!
Ads should be reviewed with the specific trademark scenarios in mind and approved when appropriate. If your ad gets disapproved and the term cited doesn't seem appropriate for trademarking within your industry, please reply to the disapproval email. Like sailorjwd found, incorrect disapprovals will be fixed.
AWA2