The company is adding new tools as bad actors use ChatGPT-themed lures and mask their infrastructure in an attempt to trick victims and elude defenders.