Subjects
Annual Updates Subcategories
Anti-Bribery & Compliance
Anti-Trust & Competition
Collective Agreements
Data Protection & Privacy Subcategories
Disputes and Grievances
Employment Law
Employment Records
Employment Status Subcategories
Equality, Diversity and Equal Rights Subcategories
EU Legislation
Industrial & Workplace Relations Subcategories
International Mobility Subcategories
Parental Leave
Pensions & Benefits
Protection of Business Interests Subcategories
Recruitment, Termination of Employment Subcategories
Remuneration Subcategories
Social Security
Social Security Contributions
Termination of Employment
Terminations Subcategories
Working Hours & Pay Subcategories
Workplace Culture Subcategories
Workplace Investigations
Jurisdictions
Americas Subcategories
APAC Subcategories
Cross Border
EMEA Subcategories
Global

USA: AI and Client-Lawyer Privilege

By - - AI Compliance, Protection of Business Interests

Do conversations with a chatbot (e.g. Claude or ChatGPT) in relation to litigation, remove the usual protection provided by client – lawyer privilege?

A judge in a New York Federal court has recently ruled that no attorney – client relationship exists “or could exist, between an AI user and a platform such as Claude”.

Privilege

Privilege in the US, and in many jurisdictions, enables lawyers and clients on litigation matters to communicate information and advice to each other, without the risk of having to disclose those communications to the opposing side or the court. The rules can be detailed and complex. To lose privilege could be very damaging to a case.

Recent Cases

The NY Case

In the New York case, which was a criminal matter relating to a federal prosecution for securities and wire fraud, the defendant had used Claude to prepare reports on his defence, to share with his attorneys. 

The defence argued that these “chats” and their resulting documents were protected by privilege. The judge disagreed and required that the resulting 31 documents be handed to prosecutors. He observed that chatbots were not lawyers and therefore privilege did not apply.

Although this was a criminal case, the same privilege rules would apply to civil cases.

Another Case

Meanwhile in Michigan, on the same day, a Magistrate Judge held that a woman representing herself in an employment claim against a former employer, did not have to hand over her conversations with Chat GPT in preparing her case.

The judge said that her AI chats were part of her personal “work product”, and in this case the AI chatbot is a “tool” and not a person.

Chatbot Confidentiality Obligations

Standard AI platform terms usually state that they can share users’ data with third parties. Claude terms expressly make clear that users have no expectation of privacy and advise users to consult a lawyer before relying on chatbot legal advice.  

How are US Attorneys Responding?

As is well known, people increasingly use AI chatbots for legal advice and for the preparation of legal documents. The question is, is it safe in litigation matters? Obviously, it is new area, and further rulings will in time clarify the position.

Meanwhile, US law firms are increasingly taking steps to protect their clients from mistakes:

  • Advising clients not to use chatbots as trusted confidents.
  • Terms of engagement increasingly warn clients that the use of chatbots could remove the confidentiality and privilege protection.  
  • Some firms advise that using “closed” AI systems may help reduce the risk of losing privilege.
  • The best advice remains “do not discuss your case with anyone (now to include AI chatbots), except your lawyer”.     

This is a high-level general update only. Legal advice should be obtained on specific circumstances.

Tags:
Scroll to Top