AI-generated docs arent covered by attorney-client privilege, judge says

0
18

Judge: AI-made docs sent to lawyers are admissible in court

AI is once again on the witness stand.

In comments made during preliminary trial proceedings, Manhattan-based U.S. District Judge Jed S. Rakoff decided that documents prepared by an AI tool and then shared with an attorney are admissible in court, existing outside of attorney-client privilege, reported Law360.

The decision relates to the case of Beneficient CEO Bradley Heppner, charged with committing $150 million in securities and wire fraud between 2018-2021. The financial services exec was indicted in November. Prior to his arrest, Heppner used Anthropic's Claude chatbot to create 31 documents later shared with his defense attorney, evidence which was then seized by investigators. Federal prosecutors say that the documents are fair game and should be regarded as a "work product," rather than a confidential legal strategy. They also say that the AI tool's own usage policies do not guarantee confidentially.

Mashable Light Speed

The defense argued that, even though they weren't created by attorneys, the documents include information based on conversations with legal representatives that should be shielded. Also, by implicating the defense team itself in the documents, the evidence could create a conflict of interest between Heppner and his own representation, they added. Rakoff said he did not see any basis for claims of attorney-client privilege, but agreed the evidence could result in a witness-advocate conflict and a mistrial.

The comments shine a spotlight on an escalating conflict among AI developers, privacy watchdogs, and safety advocates.

This Tweet is currently unavailable. It might be loading or has been removed.

AI executives, including OpenAI CEO Sam Altman, have suggested extending the same kind of attorney and therapist client privilege, which protects communications from court discovery, to conversations with AI chatbots. Altman has argued that, in particular, increasingly personal uses of AI assistants, including those designed for therapy or health advice, necessitate a reconsideration of what kinds of communications are granted legal privileges. Facing a multitude of lawsuits for copyright infringement and AI's effect on mental health and youth safety, AI developers have fought to keep chatbot conversations behind the black box, despite many providing express permission for government entities to view chat logs.

Meanwhile, chat history data has been essential in cases alleging misconduct and safety failures among AI's big players. Simultaneously, privacy watchdogs have raised concerns about extensive data collection and storage by AI tools, which has in turn prompted AI developers to institute measures to minimize chat history storage and allow users to use AI incognito.

Search
Categories
Read More
Home & Garden
I Tried Ina Garten’s Favorite Valentine’s Day Dessert, and It Was Love at First Bite
I Tried Ina Garten’s Favorite Valentine’s Day Dessert, and It Was Love at First Bite Credit:...
By Test Blogger9 2026-02-07 18:00:30 0 390
History
15 Fascinating Types of Soldiers Who Fought in the Napoleonic Wars
15 Fascinating Types of Soldiers Who Fought in the Napoleonic Wars 7....
By Test Blogger2 2026-01-28 06:00:55 0 611
Games
Hytale Update 2 adds a "temporarily bad" new grimoire, but Hypixel boss steps in to stop "Reddit going full panic mode"
Hytale Update 2 adds a "temporarily bad" new grimoire, but Hypixel boss steps in to stop "Reddit...
By Test Blogger6 2026-01-23 22:02:22 0 579
Games
Fortnite just added official Dungeons and Dragons islands, and here are the codes you need
Fortnite just added official Dungeons and Dragons islands, and here are the codes you need...
By Test Blogger6 2026-01-24 01:00:44 0 582
Technology
White House adds AI-produced tears to image of arrested protestor
White House uses AI to alter image of arrested anti-ICE protestor...
By Test Blogger7 2026-01-23 22:02:34 0 570