Tiffany Lee

Palo Alto
Tiffany Lee is the managing partner of Willkie's first West Coast office in Palo Alto, which she played a pivotal role in launching in December 2018. A dealmaker, she has provided counsel on some of the most intricate and high-stakes transactions within the heavily regulated technology sector. Her expertise particularly shines in pioneering artificial intelligence projects, navigating the complexities of an area where regulation is still catching up.
"I started taking on matters involving AI as early as 5 years ago, when we were helping one of our clients with developing and licensing semiconductor chips for its AI infrastructure," Lee said. "The idea that computers and robots can simulate human intelligence has always been fascinating to me growing up -- of course it was also greatly fueled by Hollywood with classic movies such as 'Terminator' and 'Blade Runner.'"
Lee's approach to staying informed and ahead in her field involves extensive reading and hands-on experience, particularly in AI transactions with various clients. The Andy Warhol case has been a significant motivator for her to stay abreast of copyright cases impacting AI because she said it "'muddied' the analysis by not only looking at whether the output is considered a 'transformative use' of the original work, but also by looking at how the output is used to compete with the original work.
"The Warhol case is particularly relevant because generative AI models are trained on images and texts that can include copyrighted works, so what is considered transformative use for purposes of a fair use analysis has important implications on whether generative AI models training on copyrighted works and the resulting output could be considered copyright infringement," Lee explained.
In the past few years, Lee has assisted with negotiating licensing agreements to license images to train large language foundational models and have also assisted a client with drafting the licensing terms and conditions for the commercial use of its large language foundational model.
"They were among my most significant matters because we had to balance and anticipate the risks involved in an unsettled area of law with the interests of the business team and engineers to make the large language foundational model widely adopted with 'easy to understand' terms," she said.
Lee also recognizes the significance of President Biden's Executive Order on Safe, Secure and Trustworthy Artificial Intelligence issued in 2023. She anticipates that over the next two years, there will be a surge in Federal regulation, standards, and guidance on AI usage, promoting a more coordinated and uniform approach across Federal agencies to manage the safety and security risks associated with AI.
"In addition, it is expected that at a state level, more states will adopt more comprehensive resolutions and enact legislation to govern AI, particularly in protecting the rights of IP owners and consumers," Lee said. "I anticipate seeing a continuing uptick of AI copyright cases being filed, particularly relating to the use of copyrighted works to train generative AI models and whether output from the AI models constitute copyright infringement of the original works from which it was trained."