ChatGPT Has No Legal Privilege

Sam Altman, CEO of OpenAi, maker of ChatGPT confirmed in a recent podcast interview that ChatGPT has no legal privilege. Therefore, anything you share with ChatGPT can be used as evidence.

ChatGPT has no legal privilege

In podcast episode with Theo Von, Sam explained that many users have begun treating ChatGPT like a therapist, life coach, or even an informal legal advisor. Users share deeply personal and sensitive information assuming it will remain private. But unlike communications with a solicitor or therapist, there is currently no legal privilege attached to conversations with AI models.

Implications

This has serious implications particularly for trainee and junior solicitors who may be tempted to use ChatGPT to “talk through” stressful situations or even hypothetical legal scenarios. Sam noted that AI transcripts could be subpoenaed in legal proceedings.

Sam urged lawmakers to consider creating new forms of privilege or data protection for AI interactions. But for now, users should assume that anything they type into an AI tool can be logged, stored, and disclosed.

Key Takeaways

• Never assume AI tools are private, treat all conversations with AI as if they could one day be made public.
• Avoid using AI tools to explore sensitive legal scenarios that could be interpreted as offering or seeking legal advice.
• Stick to AI for research or admin help, not emotional or strategic decision making around your practice or clients.
• Educate others in your firm. Especially non legal staff about these risks if they’re also experimenting with AI.

As AI evolves faster than regulation, staying cautious and informed is essential.

Using AI in the legal profession

Should you use AI to study for the SQE: Although AI like ChatGPT can answer some sample SQE questions correctly, it still makes mistakes. Use trusted materials and be cautious when self‑studying

AI law firm approved, AI Lawyer fined: The SRA granted landmark approval to Garfield AI, an AI‑powered small‑claims law firm, while a separate “AI lawyer” app was fined for misleading claims.