Patrick D. Curran, a partner at Quinn Emanuel specializing in AI legal issues, discusses critical topics surrounding artificial intelligence in this engaging conversation. He highlights the inadequacies of patent law in protecting AI innovations, especially around algorithms and prior art. Curran reveals how companies are increasingly leaning on trade secrets for protection, despite enforcement challenges. The conversation touches on AI's role in invention and the intricate legal battles over copyright and data scraping in rapidly evolving tech landscapes.
43:49
forum Ask episode
web_stories AI Snips
view_agenda Chapters
auto_awesome Transcript
info_circle Episode notes
insights INSIGHT
AI Model IP Protection
AI models, despite their high value, often face challenges in patent protection due to their mathematical nature.
Companies are increasingly relying on trade secret protection, despite its enforcement difficulties, to safeguard their AI innovations.
insights INSIGHT
AI-Generated Inventions and Patent Law
AI-generated inventions lack patent protection without significant human involvement, creating a gap in IP law.
This raises questions about the nature of invention and the role of humans in the process, potentially impacting patent law's relevance.
insights INSIGHT
Copyright and AI Training Data
Courts are grappling with whether using copyrighted material to train AI models constitutes fair use or infringement.
This involves reinterpreting copyright principles in the context of AI, considering factors like copying, transformation, and economic impact.
Get the Snipd Podcast app to discover more snips from this episode
John is joined by Patrick D. Curran, Partner in Quinn Emanuel’s Boston and New York offices. They discuss the emerging issues regarding artificial intelligence currently before the courts, legislatures and government regulators and that, while many critical questions are pending before courts and regulators, clear answers are still few and far between. First, they discuss how despite the billions of dollars being invested in developing large language AI models, patent law often does not protect those investments because patents generally do not cover general ideas, mathematical concepts, or algorithms. They also discuss the question of whether an AI generated invention may be cited as prior art that would invalidate a human-generated invention. Patrick then explains that companies are increasingly relying on trade secret protections to safeguard their AI innovations, even though this approach comes with challenges. Patrick further explains that trade secret protection may extend indefinitely, unlike patents which expire after a defined term, but notes the difficulty inherent in detecting when competitors might be using proprietary models, making trade secrets harder to enforce. They also discuss AI's role in invention, noting that while AI may create invent things, such as new molecules, if there is no human involvement in the process, the discovery cannot be patented. They then examine the legal challenges regarding the use of copyrighted material in training AI models, including whether using copyrighted material for AI training constitutes fair use, the degree to which companies can limit data scraping through their terms of service, and the role that technical safeguards against scraping might play in future disputes. They also discuss recent defamation claims based upon AI generated content and the difficulties of proving intent when human input to the content is minimal. The discussion then turns to recent regulatory developments, including recent legislation in US cities such as cities like New York City and Portland, Oregon, states including Colorado and California and international efforts like the European AI Act and the “Brusselization” of GDPR requirements. Patrick describes the industry's divided stance on regulation, with some companies calling for stricter oversight while others fearing that regulation will stifle innovation. Finally, both John and Patrick agree that as courts and regulators tackle these complex issues, the legal landscape surrounding AI will continue to evolve rapidly.
Law, disrupted