Sam Altman: ChatGPT Therapy Chats Offer No Legal Confidentiality

July 28, 2025Provided by Utku Ege Tuluk

OpenAI CEO Sam Altman recently issued a stark warning: conversations with ChatGPT, especially those of an emotional, personal, or therapeutic nature, do not receive the same legal confidentiality protections as interactions with human professionals—such as therapists, doctors, or lawyers. This gap means that user chats can legally be accessed and submitted as evidence in court cases.

Altman made these points during an appearance on Theo Von’s podcast, This Past Weekend with Theo Von. He noted that many users, especially younger individuals, treat ChatGPT as a virtual therapist or life coach—often sharing deeply personal or sensitive information (TechCrunch).

Unlike licensed professionals who operate under well-defined privilege laws (e.g. doctor–patient or attorney–client confidentiality), AI interactions currently fall outside such legal protections. If a lawsuit arises, OpenAI could be compelled to produce those chat logs under legal discovery rules (Business Insider).

⚠️ Key Highlights

No legal privilege: ChatGPT conversations are not shielded by confidentiality laws.
Data may be subpoenaed: In the event of lawsuits, courts could order OpenAI to hand over chat transcripts.
Missing regulatory framework: Altman emphasized the urgent need for new laws or policies—what he calls “AI privilege”—to cover these types of digital conversations (Outlook India, Business Insider).
Ongoing legal case: OpenAI is currently appealing a court order in a lawsuit brought by The New York Times, which seeks to force the retention of all user chat logs—except for certain enterprise accounts or API users with zero‑data‑retention options (TechRadar).

Altman labeled the situation “very screwed up,” arguing that AI conversations should be protected in the same way as human‑to‑professional exchanges. He also acknowledged that concerns about privacy and legal ambiguity can deter users from engaging more deeply with ChatGPT until there’s legal clarity (TechCrunch).

🔍 Why This Matters

As AI becomes increasingly integrated into daily life, especially in areas tied to emotional well‑being and relationship advice, users may feel safer opening up to chatbots. But without legal protections:

Users risk unexpected exposure of personal thoughts in legal proceedings.
Trust in AI platforms may erode if people feel their privacy isn’t guaranteed.
Ethical and legal standards for AI interactions remain undefined, raising broader concerns about data rights, security, and accountability.

📌 What You Can Do

Avoid sharing deeply personal or sensitive information via ChatGPT if legal protection matters to you.
Review your chat settings—deleting conversations removes them from your account but doesn’t guarantee total erasure or immunity from legal requests (Business Insider, The Indian Express, The Times of India).
Consider using services with strong privacy safeguards, such as licensed mental health professionals or end-to-end encrypted platforms.
Stay updated on legal developments and privacy policy changes, particularly around OpenAI’s litigation and proposed regulatory reforms.

✅ Summary Table

Topic	Key Point
Legal Status	ChatGPT conversations are not legally privileged.
Data Access	Chat logs can be subject to court orders or subpoenas.
Regulatory Gap	AI privilege laws have not yet been established.
Current Dispute	OpenAI appealing order to retain all user chats indefinitely.
User Implication	Users shouldn’t expect the same privacy protections as with professionals.

If you’re considering using ChatGPT—or any AI chatbot—for emotional support or sensitive counseling, be aware: your conversations may not be private in the eyes of the law. OpenAI’s leadership is advocating for new frameworks to protect these digital conversations—but until they materialize, caution remains advisable.