LSPN Connect: artificial intelligence and IP
Artificial Intelligence (AI) offers exciting and unprecedented opportunities for life sciences and healthcare but also poses complex questions for the IP sector, as an LSPN Connect session discovered yesterday, July 21.
Ryan Abbott, professor of law and health science at the University of Surrey, John Sninsky, translational medicine consultant and Nathan Zhang, in-house litigation counsel at Applied Materials, joined publisher and editor-in-chief of WIPR, Peter Scott, to discuss the IP challenges presented by AI.
Abbott, author of “ The Reasonable Robot: Artificial Intelligence and the Law”, explained there were significant issues in the AI patenting landscape.“The question is around how do you patent AI? Traditionally, software per se has been patentable but computer implemented inventions have not been. That [question] has been on the agenda for decades but has moved back and forth in response to jurisprudence and new policies in the US and Europe. It continues to be an ongoing discussion.”
There were also intriguing challenges around how to protect inventions when AI has played a “substantial role in the invention”, he added, drawing upon his experience of seeking a major shakeup in patent law after his research team filed applications listing a machine, Dabus, as a sole inventor.
The panel tackled the question of whether courts and patent offices will eventually view AI as an inventor, skilled in “prior art”, and dramatically upturn long-standing legal precedent. Zhang argued that this scenario would present challenges for the IP sector, given that for example, while you can depose a person, it is not possible to depose an AI.
Abbot, however, queried this and argued that the AI legal landscape could present some fascinating litigation. “You can’t depose an AI? I’m not so sure about that. I think that with forensics, you might be able to get a better answer about how an AI came up with an output. After all, people lie under oath.”
The session explored how different regulatory and legal systems in the US, Europe and China had also created problems in devising a definitive approach to AI patenting.
Abbott said: “China wants to be an AI leader but I have not seen that reflected in the changes to its legal system or its regulatory framework. I have not seen particularly progressive AI laws. China’s law has not caught up to its policies, or at least its foregn policy objectives.”
He added that it was more difficult to patent AI in Europe than the US. “Historically, there have been a lot of regulatory barriers in Europe, particularly around data privacy and ownership, consumer rights and legislation such as the EU General Data Protection Regulation (GDPR). The evidentiary bar for patents is higher in Europe,” he said.
Questions around data had become particularly controversial, added Abbot. “So many AI applications rely on training with data and machine learning, so that has taken on an added importance.” Sninsky added, however, that the US Supreme Court had “stumbled” and increased the “murkiness” around US patent law in recent decisions.
The panel also debated the merits of trade secrets compared to the patenting route, when it comes to protecting the IP around AI-led inventions.Zhang said: “It is incredibly important to emphasise that a trade secret is not static. You have to make sure it stays a secret.” He noted that this route was not a sure-fire alternative to patenting and the fast-paced nature of the AI employment space posed significant problems.
“There are so many start-ups and it is a bit of a ‘goldrush’ at the moment, especially when it comes to leveraging AI in the healthcare space. It is very difficult to keep track of people’s coming and goings, and who might divulge secrets pertaining to your project.”
The panel agreed that the trajectory of AI would transform the life sciences legal sector, and to maintain and achieve success, law firms in this space will need to offer a broader, more technical skill set. “We are going to need litigators who understand computer science to understand these answers,” said Abbott.
LSPN Connect is the membership programme for the Life Sciences—to watch on this session and for more information on joining, visit www.lspnconnect.com
Did you enjoy reading this story? Sign up to our free daily newsletters and get stories like this sent straight to your inbox.
Today's top stories
USPTO welcomes 'promising' women inventor numbers
US accuses Chinese nationals of decade-long hacking campaign
