• Pascal's Chatbot Q&As
  • Posts
  • The Lawdragon feature on lead attorneys in the New York Times and MediaNews Group lawsuits against OpenAI and Microsoft offers a revealing look into the legal frontline of AI copyright litigation.

The Lawdragon feature on lead attorneys in the New York Times and MediaNews Group lawsuits against OpenAI and Microsoft offers a revealing look into the legal frontline of AI copyright litigation.

These landmark cases raise foundational questions about fair use, intellectual property rights, and the commercialization of journalism and other media.

Author

Pascal Hetzscholdt
August 07, 2025

Suing ChatGPT—The Lawyers Leading the Fight for Copyright Protection in AI

by ChatGPT-4o

The Lawdragon feature on Steven Lieberman and Jennifer Maisel—lead attorneys in the New York Times and MediaNews Group lawsuits against OpenAI and Microsoft—offers a revealing look into the legal frontline of AI copyright litigation. These landmark cases challenge the unchecked ingestion and use of copyrighted content to train large language models (LLMs) like ChatGPT and Copilot, raising foundational questions about fair use, intellectual property rights, and the commercialization of journalism and other media.

Valuable Statements and Findings for Plaintiffs Litigating Against AI Makers

For those currently engaged in litigation against AI developers, the article provides several important insights and precedents:

1. Procedural Victories Are Achievable

“The New York Times complaint and The Daily News complaint have survived broad-ranging motions to dismiss… All of the copyright claims are intact…”
✅ Significance: Surviving a motion to dismiss is a critical milestone, affirming that courts are taking such copyright claims seriously, not dismissing them as incompatible with fair use at the pleading stage.

2. Model and Training Data Inspection Protocols Are Being Established

“We have a training data inspection protocol and step-by-step instructions… We have a model inspection protocol… We’re literally building the plane as we’re flying it…”
✅ Significance: Plaintiffs should demand structured discovery protocols—this precedent helps unlock the opaque nature of AI training and output behavior, critical for proving unauthorized use of copyrighted material.

3. The VCR Analogy is Misleading

“Microsoft and OpenAI argue it’s like selling VCRs. But that’s not the right analogy. It’s like they are selling VCRs preloaded with thousands of copyrighted movies…”
✅ Significance: This reframing rejects tech companies’ common defense that their tools are “general purpose.” Plaintiffs can use this analogy to argue that LLMs are not neutral platforms when preloaded with copyrighted data.

4. Generative Outputs Can Trigger Trademark and False Attribution Claims

“Generative AI output confabulates facts and information… and uses our client’s trademarks in a way that’s incredibly damaging.”
✅ Significance: Plaintiffs can assert claims beyond copyright—in particular, trademark dilution, brand confusion, and reputational harm when LLMs hallucinate false attributions to real publishers.

5. Existential Harm to Content Creators

“It’s an existential danger for them when somebody can just take their content and regurgitate it without any compensation.”
✅ Significance: Demonstrating economic harm is vital in court. This quote helps frame the issue as not just a legal violation, but a fundamental threat to sustainable journalism and knowledge production.

6. Building a Precedent for New Areas of Copyright Law

“We’re creating new precedent and we’re creating new law… the facts really drive the analysis.”
✅ Significance: These are not settled matters. Litigants can lean into the novelty of AI’s implications for copyright law—using factual discovery to shape new legal boundaries.

“Artificial intelligence is going to create so many new risks… she’s going to need to double the size of her department to deal with all the issues.”
✅ Significance: Plaintiffs should anticipate complex, multifaceted legal battles involving privacy, IP, consumer protection, and potential harms from unsafe AI deployment.

“I can give Steve an explanation of the technology and he’s able to learn it and use those explanations in really compelling ways.”
✅ Significance: Legal teams must combine deep IP litigation expertise with fluency in AI systems and technical evidence gathering—especially when dealing with black-box models.

Recommendations for AI Makers

To mitigate legal risk and build sustainable AI ecosystems, developers and companies should take the following steps:

1. Obtain Content Licensing Proactively

Develop AI training datasets from licensed, permissioned sources. Negotiate directly with rights holders (e.g., media companies, publishers, educational institutions) and set up frameworks akin to Spotify’s licensing model—where usage is traceable and compensable.

2. Ensure Transparency of Training Data

Maintain accurate documentation of all training sources, including clear provenance metadata. Be prepared to disclose this under protective order in litigation or regulatory inquiry.

3. Establish Output Attribution Safeguards

Prevent false attributions by building rigorous guardrails in generation layers. Outputs referencing a real publisher or individual must have factual grounding or explicit disclaimers to avoid trademark violations or defamation claims.

4. Engage in Risk Assessment and Pre-deployment Review

Create AI governance teams that collaborate with legal, compliance, and technical experts to review the intellectual property risks of each new deployment. This includes use in commercial products like Copilot or summarization tools.

5. Design Models to Respect ‘No-Crawl’ and Robots.txt Directives

Even if the law has yet to fully clarify their enforceability, respecting opt-out signals shows good faith and could help avoid litigation and reputational damage.

6. Collaborate with Standards Bodies and Regulators

Participate in the development of industry-wide protocols for ethical data sourcing, model explainability, and fair compensation structures to avoid fragmentation and legal uncertainty.

Conclusion

The article provides a blueprint for how copyright owners can build compelling AI litigation strategies and how courts may soon reshape the balance between innovation and intellectual property. For AI developers, the message is clear: act now to embed legality and legitimacy into your model development and deployment lifecycle—or face an avalanche of litigation that could shape the future of the AI economy.