Close
How GenAI services might use your data and how to keep your data safe.
MediaCentral Widget Placeholderhttps://mediacentral.ucl.ac.uk/Player/0dEa6C2d
In this video: Safeguarding your data with GenAI
GenAI systems were originally trained on publicly available data from the web and social media. This raised concerns that copyrighted materials were being exploited unfairly. Increasingly GenAI developers are making agreements with publishers to use their content under licence. This includes publishers of academic journals and books.
The way that the data you upload and the prompts you write are used will depend on the and conditions you have with a GenAI provider. Free accounts will often allow the provider to use your data to continue to train their models. Paid-for or corporate accounts may allow you to opt out of your data being used for training purposes.
UCL has an agreement with Microsoft Copilot for all staff and students. If you sign in with your UCL credentials you can be confident that your prompts, questions and results are protected, the chat data is not saved, and Microsoft does not have access to it. This means no one can view your data and your data is not used to train the AI models.
However, we would always advice against uploading any sensitive or personal data, or any data to which you do not own the copyright or have permission to use.
