AI and client confidentiality

We are aware of individuals uploading confidential information, legal advice, correspondence and draft documents into publicly available AI systems like ChatGPT, Claude, Google Gemini and others.
 

People do this for entirely understandable reasons: to prepare instructions when seeking legal advice, critique legal advice, or generate summaries to circulate throughout an organisation. However, we strongly advise clients against doing this.
 

Inputting legally sensitive information into publicly available AI platforms jeopardises confidentiality and legal professional privilege.
 

An AI-generated summary is a document produced through the interaction of the user and an AI system – it is not the lawyer's communication, and it is therefore unlikely that privilege extends to that document.
 

If privilege is lost or successfully challenged, communications and documents that clients assumed were protected may become disclosable in court or tribunal proceedings. In addition, legally sensitive information that is entered into publicly available AI systems could potentially be disclosed to a third party.
 

This guidance does not constitute legal advice and should not be relied upon as such.