AI LICENSING, CHATGPT RESPONDS TO NYT, AND EU INQUIRY INTO VIRTUAL WORLDS

Below is this week's tracker of the latest legal and regulatory developments in the United States and in the EU. Sign up here to ensure you do not miss an update.

AI Intellectual Property Update:

  • In a submission to the United Kingdom's House of Lords, OpenAI has claimed that it would be "impossible" to create AI tools like ChatGPT without access to copyrighted material:
    • "Because copyright today covers virtually every sort of human expression – including blogposts, photographs, forum posts, scraps of software code, and government documents – it would be impossible to train today's leading AI models without using copyrighted materials.... Limiting training data to public domain books and drawings created more than a century ago...would not provide AI systems that meet the needs of today's citizens."
    • The company also noted that it plans to continue "to develop additional mechanisms to empower rightsholders to opt-out of training" though it believes that "legally copyright law does not forbid training."
  • AI Licensing:
    • OpenAI is reportedly in talks with "dozens" of publishers to license content to train its AI tools. The offers from OpenAI are allegedly around $1-5 million per year.
    • Bloomberg reports that OpenAI is specifically in talks with CNN, Fox Corp. and Time to license their content. This would involve using this content to train ChatGPT. "CNN and Fox are negotiating not just around licensing text, but also video and image content."
  • Getty Images has launched Generative AI by iStock, a text-to-image platform specifically designed to make stock photos.
    • The tool was trained on Nvidia's Picasso model using only Getty's creative library and iStock's stock photo library. It did not train on Getty's editorial image library to prevent it from generating trademarks or known personalities.
    • Users of iStock will receive Getty's "standard" legal indemnification of up to $10,000.

AI Litigation Update:

  • Another copyright lawsuit has been filed against OpenAI and Microsoft. Nicholas Basbanes and Nicholas Gage, two nonfiction authors and investigative journalists, have filed a proposed class action complaint (S.D.N.Y. No. 1:24-cv-00084) alleging direct, vicarious, and contributory copyright infringement in OpenAI's use of their copyrighted works as training materials.
  • OpenAI has responded to the New York Times lawsuit in a blog post.
    • OpenAI claims that the Times "intentionally manipulated prompts, often including lengthy excerpts of articles, in order to get our model to regurgitate" the newspaper's articles verbatim. "Even when using such prompts, our models don't typically behave the way The New York Times insinuates, which suggests they either instructed the model to regurgitate or cherry-picked their examples from many attempts," it says in the blog. "Memorization is a rare failure of the learning process that we are continually making progress on[.]"
    • "Training AI models using publicly available internet materials is fair use, as supported by long-standing and widely accepted precedents. We view this principle as fair to creators, necessary for innovators, and critical for US competitiveness."

AI Policy Update—Federal:

  • The FTC is hosting a Tech Summit on January 25 that will include panels on "AI Chips and Cloud," "AI Data and Models," and "AI & Consumer Applications." Chair Lina Khan will give opening remarks.
  • The Senate Judiciary held a hearing on January 10 about AI and journalism. Witnesses included:
    • Jeff Jarvis: Professor of Journalism, CUNY
    • Danielle Coffey: President and CEO, News Media Alliance
    • Curtis LeGeyt: President and CEO, National Association of Broadcasters
    • Roger Lynch: CEO, Condé Nast

AI Policy Update—European Union:

  • The European Commission put out a call for contributions regarding competition in virtual worlds and generative AI, asking for views on "how competition law can help ensure that these new markets [of virtual worlds and generative AI] remain competitive." The Commission may organize a workshop with relevant stakeholders to discuss these issues in response to the consultation. Among other things, the questions include topics on barriers to entry, future competition, monetization, and vertical integration. Responses are due by March 11.
    • The European Commission is also "looking into some of the agreements that have been concluded between large digital market players and generative AI developers and providers [and] is investigating the impact of these partnerships on market dynamics."
    • The EC further states that it is "checking whether Microsoft's investment in OpenAI might be reviewable under the EU Merger Regulation."
  • OpenAI, listed as the controller for the personal data of users in the European Economic Area (EEA) its Dublin-based subsidiary, effective February 15, 2023. This change entails, among others, that any concerns regarding ChatGPT's compliance with the European Union's General Data Protection Regulation (GDPR) must be referred to Ireland's Data Protection Commission rather than each individual member state.
  • Euractiv reports that the European Union will not support an exemption of the private sector from the scope of the future Council of Europe Convention on AI, i.e. the first international treaty on AI, whereas it will push for as much alignment as possible with the EU AI Act.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.