A federal judge affirmed an order requiring OpenAI to produce about 20 million de-identified ChatGPT conversation logs to The New York Times and other publishers in copyright litigation. The order covers a limited sample of retained conversations for use in controlled legal discovery—not every user’s chats and not a public release. It does not decide whether OpenAI infringed copyright.
What OpenAI has to produce
The production concerns consumer ChatGPT output logs: records of interactions that can include a user’s prompts and ChatGPT’s responses. The court ordered OpenAI to provide a defined sample of approximately 20 million retained conversations after de-identification, under the litigation’s protective measures.
As an Amazon Associate I earn from qualifying purchases.
Magistrate Judge Ona Wang ordered production, then denied reconsideration on December 2, 2025, directing OpenAI to produce the logs within seven days after de-identification. She denied a request to pause production on December 9. On January 5, 2026, District Judge Sidney Stein rejected OpenAI’s objections and affirmed the discovery orders. The orders are available in the December 2 order, the December 9 order and the January 5 ruling.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsThis is discovery: the exchange of potentially relevant evidence among parties in a lawsuit. It is not an order to publish the chats online or make them available to the public.
#1 Best Overall
- Perfect Fit: Designed for your iPhone 15 6.1 inch. These protectors are Slightly smaller than the screen to avoid interfering with Face ID or the front camera, ensuring case compatibility and preventing bubbling.
- True 28° Anti-Spy: The screen is visible only to the immediate user directly in front of it, providing effective privacy protection: 28° for partial peep-proof and 45° for complete peep-proof coverage. Perfect for public places like buses or elevators.
- Durable Protection: This privacy screen protector features 9H hardness tempered glass to guard against scratches, drops, and bumps. Its hydrophobic and oleophobic coating resists fingerprints and smudges.
- HD Clarity & Touch Sensitivity: This privacy glass screen protector maintains screen brightness and detail while ensuring smooth, accurate touchscreen response with minimal distortion.
- Bubble-Free & Anti-Fingerprint: The UNBREAKcable screen protector uses auto-adsorption to eliminate bubbles and a plasma oleophobic coating to resist fingerprints and smudges. Enjoy a crystal-clear, easy-to-clean screen.
Why the publishers sought conversation logs
The Times and other publishers contend that ChatGPT can reproduce or closely recall protected articles. They say conversation records may help show whether, when and under what prompts the system produced material resembling their works. Logs may also bear on the publishers’ infringement theories and OpenAI’s defenses, including questions about the systems’ ability to identify or search for publisher material.
The publishers’ claims are allegations to be tested in the case. A particular output’s legal significance depends on the work, prompt, response, similarity and other facts; a log that contains similar text does not by itself establish infringement.
Rank #2
![EGV Compatible for iPhone 16/iPhone 15 Privacy Screen Protector [9H+ Military Grade Shatterproof & 100% Anti Spy], Privacy Screen iPhone 16/iPhone 15 Tempered Glass, Full Coverage Protection, 2 Packs](https://m.media-amazon.com/images/I/51FepTt9CLL._SL160_.jpg)
- 【𝟐𝟓°𝐒𝐮𝐩𝐞𝐫𝐢𝐨𝐫 𝐏𝐫𝐢𝐯𝐚𝐜𝐲 𝐏𝐫𝐨𝐭𝐞𝐜𝐭𝐢𝐨𝐧】The EGV iphone 16 pro max privacy screen protector utilizes micron-level micro-louver technology to achieve a 25° anti-spy angle on both sides. This ensures maximum privacy by blocking screen visibility beyond 25°, preventing prying eyes in public spaces. Your screen stays confidential—anytime, anywhere.
- 【𝐎𝐧𝐞 𝐒𝐭𝐞𝐩! 𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐯𝐞 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐜 𝐝𝐮𝐬𝐭 𝐫𝐞𝐦𝐨𝐯𝐚𝐥 𝐬𝐭𝐢𝐜𝐤𝐞𝐫】EGV privacy screen iphone 16 pro max features advanced self-alignment, dust removal, and vacuum adhesion technology. Achieve zero dust, zero bubbles, and perfect alignment in seconds—even for beginners. With a 99.99% success rate, anyone can install like a pro.
- 【𝐌𝐢𝐥𝐢𝐭𝐚𝐫𝐲-𝐆𝐫𝐚𝐝𝐞 𝐃𝐫𝐨𝐩 𝐑𝐞𝐬𝐢𝐬𝐭𝐚𝐧𝐜𝐞】Engineered with 9H+ tempered glass iphone 16 pro max privacy screen hardened via dual-ion exchange technology, EGV protector delivers military-grade durability. It withstands scratches, accidental drops, and daily wear while safeguarding your device from impacts.
- 【𝐔𝐥𝐭𝐫𝐚 𝐇𝐃 𝐂𝐥𝐚𝐫𝐢𝐭𝐲 & 𝐒𝐢𝐥𝐤-𝐋𝐢𝐤𝐞 𝐒𝐦𝐨𝐨𝐭𝐡𝐧𝐞𝐬𝐬】Premium anti-spy materials ensure industry-leading clarity, paired with a nano-oleophobic coating that repels fingerprints, sweat, and oils. The hydrophobic surface keeps your iphone 16 pro max privacy screen pristine and delivers a buttery-smooth touch that mimics your phone’s original feel.
- 【𝐅𝐮𝐥𝐥-𝐒𝐜𝐫𝐞𝐞𝐧 𝐂𝐨𝐯𝐞𝐫𝐚𝐠𝐞】Perfectly crafted for iPhone 16 Pro Max, EGV privacy protector delivers a seamless edge-to-edge fit that works flawlessly with any phone case. This product includes: 2packs iphone 16 pro max screen protector privacy. (Note: 1 privacy protector pre-loaded in the applicator + 1 individually packaged privacy protector.)
Why OpenAI objected—and why the judge rejected its proposal
OpenAI argued that producing millions of complete conversations posed privacy risks and that most of the sample would not be relevant. It proposed searching for conversations that matched terms associated with the publishers’ works and producing a smaller set. In its account of the dispute, OpenAI says the Times initially sought about 1.4 billion private ChatGPT conversations in May 2025. That figure describes the demand as OpenAI characterizes it; it is not the court-ordered production.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Judge Stein concluded that the lower court had adequately addressed privacy by limiting the production to a 20-million-log sample, requiring de-identification and applying the existing protective order. He also said OpenAI had identified no controlling law requiring the court to choose the least burdensome discovery method. The court therefore rejected the proposal to substitute keyword-selected records for the defined sample.
Rank #3
- Perfect Fit for iPhone 16/15：Engineered exclusively for iPhone 16/15 with seamless edge-to-edge coverage, ensuring precise alignment and reliable full-screen protection.
- Advanced Privacy Protection：Features a 28°privacy filter with smooth 2.5D curved edges, preventing side glances in public. Your screen remains visible only to you—ideal for commuting, traveling, and crowded environments.
- Effortless Installation：Equipped with an auto dust-elimination tool that delivers a fast, accurate, and bubble-free application, keeping your screen perfectly clear with minimal effort.
- Military-Grade Protection：Made of nano-reinforced 9H tempered glass, SGS certified. Provides 5X stronger scratch resistance and proven durability, withstanding thousands of pressure and impact tests.
- Smudge & Fingerprint Resistant：Hydrophobic and oleophobic coating repels fingerprints, sweat, and oil—ensuring your screen stays clean, clear, and smooth to the touch.
What this means for users’ privacy
The orders provide for a protected litigation production, not public access. De-identification means removing identifying information through OpenAI’s process; it is not a guarantee that a conversation can never be linked to someone. A chat may contain distinctive locations, health details, work information or personal experiences that could reveal clues even without a name. The parties disagree about whether the protections sufficiently address the risks.
A court protective order governs how material may be handled in the litigation. It is distinct from a user’s ordinary privacy settings and should not be read as a general change to ChatGPT’s privacy controls. Nor do the orders establish that users’ identities were exposed.
Rank #4
- 【Perfectly Compatible with Moto G Power（2026）】Custom-made for the Moto G Power（2026）, this set includes 2 pieces of tempered glass privacy screen protectors, balancing protection and convenience. Note: Please confirm your phone model is for Moto G Power（2026） before purchase.
- 【Effective Privacy Protection, Preventing Information Leakage】 Adopts one-way light transmission technology—screen content is only visible to users directly in front, and appears black when viewed from angles of 30° or more on the left and right, blocking prying eyes. It fully safeguards your privacy when handling information or checking private content in public places.
- 【Compatible with Most Phone Cases, All-Round Protection】 With optimized edge design, it is compatible with most dedicated cases for Moto G Power（2026） (such as ultra-thin, anti-drop, and silicone cases), avoiding case-induced edge lifting of the protector. It achieves "protector + case" dual protection without affecting the phone's appearance and grip.
- 【Easy Installation for Beginners】 simply align , then place to achieve precise fitting, no bubbles or misalignment. The whole process takes less than 1 minute.
- 【HD Touch,Quality Assurance】 Made of 9H hardness glass, it resists drops and scratches; with 99.9% light transmittance, it restores the original HD screen quality. The protector’s surface is optimized for touch, ensuring responsive operation, which balances protection and user experience.
The production is limited to retained logs in the defined sample. The order does not establish that deleted conversations or chats never retained by OpenAI were included. OpenAI says deleted chats are generally scheduled for permanent deletion within 30 days, subject to legal, security and other exceptions; that public description should not be treated as a rule that resolves how every account, product or litigation hold is handled. See OpenAI’s explanation of its data-preservation and privacy position.
Recommended Free Tools
How the preservation order led to the production dispute
On May 13, 2025, Judge Wang ordered OpenAI to preserve and segregate output-log data that otherwise might be deleted, including data potentially subject to a user deletion request or privacy-law obligation. The order was prospective: it sought to keep potentially relevant evidence from disappearing while the case proceeded. It was separate from the later order to produce the 20-million-log sample. The preservation order explains that part of the dispute.
What happened after the January ruling
On July 9, 2026, the News Plaintiffs asked the court to sanction OpenAI. Their motion alleges that OpenAI withheld or destroyed relevant evidence, made logs unavailable or unsearchable, misrepresented its ability to search training data and output logs, and prolonged discovery. These are the publishers’ allegations, not findings by the court. A scheduling order set August 13, 2026, as the deadline for OpenAI to oppose the motion; the materials cited here do not establish a later ruling. The briefing schedule and reports from Bloomberg Law and the Associated Press describe the motion.
Quick Recap
What the log order does—and does not—decide
- It decides a discovery dispute: OpenAI must produce the defined, de-identified sample under litigation protections.
- It does not decide copyright liability: The order does not establish that OpenAI infringed the Times’ copyrights or that training on copyrighted works is fair use.
- It does not decide the merits of particular outputs: The court has not ruled that any specific ChatGPT response is substantially similar to an article or legally infringing.
- It is not a final judgment: The broader case remains unresolved. In an April 4, 2025 opinion, the court allowed major portions of the Times’ claims to proceed while dismissing or limiting some claims; that procedural ruling was not a finding of liability. See the court’s opinion.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
