Yes, for business use. Switch off model training on both platforms today. It takes about two minutes and is the most important immediate privacy change.
- ChatGPT: Settings, Data controls, turn off “Improve the model for everyone”.
- Claude: Settings, Privacy, turn off “Help Improve our AI models”.
Your chats are not automatically published or shown to other users. Turning training off adds a further layer of privacy for customer information, financial data and confidential documents. It stops future training on your chats (and Claude Code sessions) but cannot undo training that has already happened.
Then review shared conversation links, connected apps and integrations, enable multi-factor authentication, use Temporary Chat (ChatGPT) or Incognito chats (Claude) for sensitive work, and never enter passwords, API keys or recovery codes. ChatGPT Business and Enterprise and Claude commercial plans do not use your inputs and outputs for model training by default, subject to exceptions such as optional data sharing or feedback.
Related pages
- ChatGPT and Claude privacy settings for business: the full guide
- Is ChatGPT safe to use at work?
- Shadow AI check
- Browse all ABCOM FAQs
Want this looked after for you?
ABCOM has supported Sussex businesses and charities since 1996 and is an NCSC Assured Service Provider – Cyber Advisor (Cyber Essentials). Talk to a real person about your IT.
Also read
How ABCOM can help
Talk to a real person at our Burgess Hill office. We are an NCSC Assured Service Provider and have supported Sussex businesses since 1996.