Jordan R. Jaffe

San Francisco
With an undergraduate degree in computer science, Jaffe has a foundational understanding of AI concepts, which has been beneficial in his legal practice. Notably, in 2017, Jaffe represented Waymo, a self-driving car service by Alphabet, in a high-profile trial against Uber.
"That was a trade secret and patent action, and I had to become deeply knowledgeable about the technology at issue," Jaffe said. "One aspect of the case involved Waymo software that plans the car's route -- software that employed AI and machine learning. At that point, it became clear how important artificial intelligence would be as an IP and litigation topic not in the future -- but in the present."
Having further developed his AI presence since then, Jaffe currently advises AI companies on best practices for creating and using large language models and related intellectual property issues. His enjoyment of the AI field stems from its long history and its resurgence as a technology that is increasingly integrated into daily life.
In a case that he recently filed in Delaware federal court, Jaffe is serving as lead counsel to Guardant Health as plaintiff in a dispute that reflects the increasing role of AI in processing medical data. Guardant Health Inc. v. Tempus AI, 24-cv-00687 (D. Del., filed June 11, 2024).
"Guardant has developed blood tests to screen patients for multiple cancers and diseases at once," he said. "Our suit asserts that Tempus infringed five of Guardant's cancer diagnostics patents in creating its own copycat diagnostic DNA tests."
He also serves as lead counsel for Wiz, a cloud security company, in a patent infringement action brought by Orca Security.
Wiz has filed a counterclaim asserting that Orca has copied its AI technology, including cybersecurity of AI models stored in the cloud and the use of large language models to enhance cybersecurity.
"One issue that comes up in AI matters involving large language models is understanding the relationship between training data, fine tuning and outputs, which is not always simple or clear cut," Jaffe said. "LLMs can be in some sense inscrutable because of this. Some commentators have even described them as like humans in this way -- it is not always easy to determine what motivated a person to say a particular statement at a particular time."