Manage risks of AI usage in corporate environments

With the recent deluge of GPT and chatGPT enabled products recently, I am wondering how the regulated industries are managing their usage given the multitude of risks associated with them. - Copyright - Model risks - Data and Privacy risks etc.

Also the embedding of these functionalities within the corporate products (SaaS) offerings is becoming a major issue, One example being Outlook starting to ingest confidential content and potentially making it available to employees that may not be purview to such info. Our organization has blocked the direct access to it, but given that its becoming so pervasive what are others doing in this scenario. The analogy to this that I can think of would be similar to enabling safe cloud usage through use of CASB's like NetSkope, but not sure if any such tech for AI usage exists at the moment.