Why Judges Are Banning Public AI in Court – Privacy Warning for Everyone
What’s happening?
In recent months, a growing number of judges have ordered lawyers and litigants not to use public artificial intelligence tools—such as ChatGPT, AI notetakers, and other open‑web chatbots—during the discovery phase of legal cases. Bloomberg Law reported in June 2026 that these rulings zero in on a specific concern: privacy and the risk of leaking confidential information.
The bans typically prohibit the use of any AI tool that sends data to a third‑party server, especially if that data might be used to train future models. In some instances, judges have also flagged the use of AI notetaking apps that record and transcribe conversations, citing potential violations of wiretapping laws.
Why the concern? Privacy and wiretapping risks
The core issue is that public AI platforms are not designed to handle sensitive or privileged information safely. When a user enters text into a free chatbot or a cloud‑based notetaker, that data often travels through the provider’s servers. Terms of service may grant the provider broad rights to use the data for product improvement, model training, or other purposes. For a lawyer handling a deposition or a judge reviewing confidential filings, that arrangement is unacceptable.
Separately, AI notetaking tools that listen in on meetings and transcribe them – even when used by one party – can run afoul of state and federal wiretapping statutes. Many U.S. states require all‑party consent to record a conversation, and an AI bot listening without everyone’s knowledge may violate that law. Bloomberg Law noted this specific pitfall in a January 2026 article, calling AI notetaking “a discovery and ethical minefield.”
Real rulings underline the risks
While many orders are sealed or unpublished, a pattern has emerged. In at least two federal district courts, judges have issued standing orders that explicitly ban the use of generative AI tools for any case‑related work unless the tool is an enterprise‑grade version that guarantees data segregation and offers terms of service that protect client confidentiality. Some courts have also instructed law firms to certify that no public AI was used in drafting documents or reviewing discovery materials.
These rulings are not about rejecting technology. They are about ensuring that the legal process’s confidentiality safeguards are not undermined by tools whose business models depend on harvesting user input.
What this means for anyone using AI
You don’t have to be a lawyer to learn from these bans. The same privacy risks apply whenever you paste personal information, financial details, medical history, or trade secrets into a public AI tool. A few real‑world examples:
- A consumer using a free chatbot to help write a sensitive email or a legal letter may be handing that text to the AI provider for use in ways they didn’t anticipate.
- Anyone who uses an AI notetaker to record a private conversation – with or without others’ knowledge – could be breaking wiretapping laws, depending on the jurisdiction.
- Small businesses that rely on public AI to draft contracts or financial plans might inadvertently expose proprietary information to a third party.
The lesson is simple: treat every public AI tool as if everything you enter could be saved, reviewed, and potentially reused.
Steps to protect your data
If you or your workplace uses AI for tasks involving sensitive information, consider these practical measures:
Use enterprise or secured versions. Many AI providers offer paid tiers that include data‑use restrictions, encryption, and promises not to train on your inputs. For legal or medical work, these are the only appropriate choice.
Assume everything you type is visible. Before pasting anything confidential into a chatbot, ask yourself: would I be comfortable if this appeared on a public website or in a competitor’s hands? If not, don’t enter it.
Review terms of service carefully. Look for language about “training,” “improving services,” or “sharing with third parties.” If the provider doesn’t explicitly commit to not using your data, treat the tool as unsafe for confidential information.
For note‑taking, get consent. If you use an AI transcription tool during meetings, inform all participants and obtain explicit consent. Check local laws – many require two‑party or all‑party consent for recording.
Ask your IT or legal team. If you handle sensitive data at work, find out whether your organisation has policies about AI use. Many companies are now issuing guidelines based on exactly these court rulings.
What to watch for next
The judges’ bans are not limited to courtrooms. They signal a broader regulatory shift: expectations about data privacy are catching up with the rapid adoption of AI. We’ll likely see more professional standards bodies, health‑care regulators, and financial oversight agencies issuing similar prohibitions. For everyday users, the takeaway is clear – understand what your AI tools do with your data, and don’t assume that convenience is worth the privacy cost.
Sources
- Bloomberg Law News, “Judges’ Public AI Bans During Discovery Zero in on Privacy Risk” (June 5, 2026).
- Bloomberg Law News, “AI Notetaking Poses Wiretapping, Discovery, and Ethical Pitfalls” (January 28, 2026).