Microsoft built supercomputer to help OpenAI infringe copyrights, NYT alleged

2026-06-26 · Source: AI - Ars Technica · Field: Legal & Regulatory — Intellectual Property & Patents, Compliance & Risk Management · Depth: Intermediate, short

Summary

The New York Times (NYT) has proposed amending its copyright complaint against OpenAI and Microsoft, alleging Microsoft actively encouraged infringement by building a bespoke supercomputing system. This move follows a Supreme Court ruling that established a new standard for contributory infringement, requiring plaintiffs to prove intentional inducement. The NYT seeks to align its claim against Microsoft with this precedent, specifying the supercomputer was tailor-made to train AI on copyrighted works, disproportionately weighting NYT articles. The original 2023 lawsuit claimed ChatGPT illegally trained on NYT content, produced verbatim outputs, caused market harms by substituting subscriptions and losing affiliate commissions, and resulted in reputational damage from false attributions. The NYT now alleges Microsoft's deployment of "Times-trained LLMs" boosted its market capitalization by a trillion dollars. Evidence includes ChatGPT outputs that skirt paywalls and generate near-verbatim excerpts, alongside hallucinations falsely citing NYT content. OpenAI maintains its training constitutes fair use, but the NYT's focus on market harms could challenge this defense, potentially leading to model wipes and extensive damages.

Key takeaway

For Directors of AI/ML or AI Product Managers building and deploying large language models, this amended complaint signals heightened legal scrutiny on training data provenance and model output behavior. You must meticulously document your data sourcing, especially for proprietary content, and rigorously audit model outputs for verbatim reproduction or content that could substitute for original works. Proving intentional inducement for copyright infringement is now a key legal standard, making transparent development practices and robust IP compliance critical to mitigate significant legal and financial risks.

Key insights

New legal precedent requires plaintiffs to prove intentional inducement for contributory copyright infringement in AI training cases.

Principles

• Contributory infringement requires proving intent.
• AI training on copyrighted works faces market harm scrutiny.
• Tailor-made systems can imply intent to infringe.

In practice

• Document AI training data sourcing.
• Analyze model outputs for verbatim content.
• Assess AI's market substitution potential.

Topics

• Copyright Infringement
• AI Training Data
• Large Language Models
• Contributory Infringement
• OpenAI Litigation
• Microsoft Supercomputer

Best for: CTO, Executive, VP of Engineering/Data, Legal Professional, Director of AI/ML, AI Product Manager

Related on AIssential

• Encyclopedia Britannica sues OpenAI for training on nearly 100,000 articles without permission — AI models' ability to reproduce copyrighted content from training data is a central legal and technical challenge.
• Disney + Sora: Billion Dollar Partnership Signals New Era — Major IP holders are shifting AI training from open access to high-value licensing deals, using litigation as leverage.
• Penguin v. OpenAI: AI-assisted production of market substitutes in the very formats that already plague children’s publishing (fake titles, lookalike covers, rapid self-publishing). — AI models reproducing copyrighted content on demand can constitute actionable copyright infringement.
• Intellectual Property vs. Machine Learning: Data Ownership and AI Models in the Age of Litigation — AI's reliance on vast datasets without clear consent creates profound, unresolved intellectual property and ownership disputes.
• Judge Rules Blacked.com Can Sue Meta for Scraping Its Porn — Copyright infringement occurs upon unauthorized downloading and distribution, regardless of AI training intent.
• Kleiner v. Adobe is another step in a pattern: the legal system is increasingly treating “training data governance” as a compliance domain, not a research footnote. — AI training data governance is evolving into a critical legal compliance domain, not merely a research consideration.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.