Forgot your password?
typodupeerror
AI Privacy The Courts

Chatbot Conversations Can Become Evidence in Court (yahoo.com) 22

The Washington Post looks "a growing number of Americans who have had their private conversations with AI chatbots made public in court," calling the chatbots "custodians of a fast-growing trove of intimate data that can be sought by law enforcement, or companies or other opponents in civil litigation." A Washington Post review of public records and local news stories found that chatbot logs were cited in 12 court cases over the past two years. It's hard to know how often chatbot material is drawn into investigations and legal proceedings more broadly, because police, and parties in civil cases, don't have to present in court all the evidence they obtain... In some cases, AI companies have reported people on their platforms to the FBI, leading to prosecutions and at least one conviction... OpenAI has said the company uses software to scan conversations for signs of dangerous behavior and flag them to human reviewers.

If a reviewer determines a chat indicates "an imminent and credible risk of harm to others," they report the user to law enforcement, the company said in an April blog post. A spokesperson for OpenAI declined to disclose how many times the company had reported users to police.... Data released by OpenAI shows requests by government agencies and law enforcement for user data growing rapidly. In the second half of 2025, the company disclosed data from more than 80 accounts, an increase of more than four times as many over the same period the year before...

When police ask someone to open their phone during questioning or after an arrest, they often do so, allowing officers to scroll through the apps they use, [according to Michael Price, the litigation director for the Fourth Amendment Center at the National Association of Criminal Defense Lawyers]... In general, the Fourth Amendment, which protects people against unreasonable searches and seizures of their property, provides the right to refuse a search of the contents of their phone unless police have a warrant. "The short answer is that most people consent," Price said. "They shouldn't, but they do..."

OpenAI CEO Sam Altman has argued that conversations with chatbots should receive special legal protections, similar to conversations between clients and their attorneys.

Special legal protections aren't recognized by courts today, the article poitns out. A financial executive who became the subject of a federal criminal investigation for fraud asked Claude about potential defense strategies, then asked a judge to disallow the conversations from the prosecution's case, according to the article. "But the judge said that protections for conversations between clients and their attorneys did not apply: Claude is not a lawyer, and the executive's attorneys did not ask him to talk to the chatbot."

The article includes quotes from three ChatGPT conversations that later turned up in court proceedings:
  • "I deleted a email a year ago from my yahoo account. Can yahoo still retrieve the email?"
  • "How f----d am i bro. What if i smahsed the s--- oitta mutlipls cars."
  • "My dad Said that I'm will get a settlement worth of 1million dollar."

"As AI becomes a bigger part of everyday life, its use as evidence and in trials is likely to become more prevalent," the article points out — especially with the growing popularity of agents that take actions for users.


Chatbot Conversations Can Become Evidence in Court

Comments Filter:
  • Seriously, Sam? (Score:5, Insightful)

    by Brain-Fu ( 1274756 ) on Saturday August 29, 2026 @03:15PM (#66313100) Homepage Journal

    Saying that conversations with a chatbot should receive the same legal protections as conversations with a lawyer is beyond ridiculous. None of the important reasons why conversations with a lawyer get such protections apply.

    Yes, it's true that some people really open up to chat bots. That still doesn't justify such protections. People just should know better than to have dangerous conversations with computer programs hosted on some one else's machine! People who really need people to talk to should still abstain from confessing their crimes to chatbots.

    Everything else we do must be done cautiously. Alcohol use doesn't come with special legal protections; it's on us to drink responsibly. Gun use doesn't come with special legal protections it's on us to manage them responsibly. Conversations with our employers, family members, friends, lovers, etc., don't come with special legal protections; it's on us to keep our mouths shut about certain topics. Why in the world would chatbot use be any different? Seriously, this is nonsense.

    People with money can buy a beefy machine and host a chatbot model on it locally. Everyone else can just develop some AI street smarts. Or go to jail after they confess crimes to a chatbot.

    • by Anonymous Coward

      Conversations with our employers, family members, friends, lovers, etc., don't come with special legal protections;

      Umm. Better check that employers category again. If you start sniffing around people with serious security clearances, even if there is no classified content involved, counter intelligence will be very interested.

      It's best for even LE to avoid this. Because if it looks like a possible threat, even they might get one of those "all expense paid" extraordinary rendition flights.

      • by tlhIngan ( 30335 )

        Umm. Better check that employers category again. If you start sniffing around people with serious security clearances, even if there is no classified content involved, counter intelligence will be very interested.

        But still not legally protected. That content could easily be sealed evidence used just for the trial but it certainly is not privileged.

        Legal protections apply to attorneys and their clients, as well as doctors and patient, and to a very limited extent, married spouses.

        And even that protection is

    • by gweihir ( 88907 )

      Indeed. But the LLM cultists think they LLMs are magic. And obviously, LLM conversations will have the highest levels of protection, surely.

      In actual reality, a single simple question to any of the free LLMs or even the free search-integrated ones would have shown these people actual reality. But when you do magical thinking, you do not verify against reality, no matter how simple that would be.

    • Gee, I dunno. Maybe we should let them take the excuse "The computer did it" and just let them run with it. We might be able to eliminate the "dog ate my homework" gene altogether, which I can't really see any downsides to, after the whole Eternal September thing.
  • by larryjoe ( 135075 ) on Saturday August 29, 2026 @03:34PM (#66313114)

    A chatbot conversation that police can legally see is just as admissible in court as a private journal entry on a desk in one's home that police can legally see. That seems obvious. However, there are separate issues involved. Was the evidence gathered legally? Was there a legal warrant or request? Was there duress? Also, using a chatbot conversation in court is far different than constant review to determine if police or supervisors should be notified.

  • Why did anyone think that conversations with chatbots would be protected by law?
    • by Todd Knarr ( 15451 ) on Saturday August 29, 2026 @04:15PM (#66313160) Homepage

      They don't. They think of them as something that happens on their computer, with only them involved. They know, in the abstract, that there's a server somewhere, but the implications of that don't really sink in. The same problems applied to e-mail since forever, but everyone dismissed it as something that didn't matter since they didn't use email. Now, though, we're seeing the result of the third-party doctrine starting to hit home in ways the average person does need to care about thanks to the ubiquity of the cloud. Chatbot logs, search histories, email records, Flock cameras, Ring doorbell cameras, ALPRs, documents stored in cloud services, social media histories, messaging service histories... everything involving records stored long-term in the hands of third parties is starting to affect ordinary people in ways they can't ignore. I suspect at some point the law is going to have to change.

      • I suspect at some point the law is going to have to change.

        The law changed just a quarter century ago with section 215 of the patriot act.

    • I think these people didn't realize that it's all recorded.

      • Everything is being recorded nowadays. Either by pervert glasses, in-house cameras, or more recently flock. Just be sure to have an NDA / legal contract shielding you from any kind of prosecution, turning recording business subjects, a felony.
      • by gweihir ( 88907 )

        They could have asked a simple question to any of the free LLMs out there and they would have known. Instead they assumed without any attempt to find out. Does not get much dumber than this.

        On the plus side, if this is the typical LLM user, the LLM brain-rot that is currently found in numerous studies will not destroy anything of value.

    • by gweihir ( 88907 )

      Probably more of the magical thinking so prevalent with the LLM fans. I mean a simple question to any free LLM will likely give you a very clear answer. The problem is these morons do not even ask the question. They just think LLMs are magic and great and cannot have any bad properties. I would say it could be LLM induced brain-rot, except that these people probably came in this exceptionally dumb.

  • I'm glad that the source poitned this out.
  • If you need a confessional that won't record, put the Eliza [archive.org] AI chat bot on a bootable USB stick and have fun.

  • Are the LLM-fans really this stupid and disconnected? I mean just asking the free DDG artificial idiot gives you a very clear and short answer. These people are too dumb to ask obvious questions.

  • A users own PC, or a bit less phone, holding an image file can be considered untampered and produced by the user. But a text line on a service with who knows how many employees having write access in the whole chain of systems handling the text line? I doubt that can be used as crime evidence.
  • If the conversation between a person and AI is not privileged conversation, can prosecutors and defense attorney subpoena any AI conversation a judge might have had with AI during the case?

If what they've been doing hasn't solved the problem, tell them to do something else. -- Gerald Weinberg, "The Secrets of Consulting"

Working...