The lack of consensus on patent licensing and the disconnect between IP departments and other business units remain barriers for businesses, according to panellists at the Licensing Executives Society International’s 2019 Annual Conference in Yokohama on May 27.
Looking at the standard essential patent licensing challenges in the 5G and internet of things (IoT) era, Dylan Lee, deputy director, licensing and IP transaction division at the IP department at Huawei in China, said that a large number of firms will implement mobile communication standards, but many of these will be SMEs. Bilateral licensing is hard given the number of implementers, he said, while connectivity needs and prices in the IoT will vary. Given the number of implementers, litigation risks will grow, he added.
Lee said that the declared aggregate royalty rates currently applied in 4G are too high and will need to be lowered in the 5G era, given that more devices will be connected with more users. A one-size fits all approach is not desirable. Although patent pools are an option, they have been largely unsuccessful in the smartphone industry. However, Lee said that they can be possible if licence terms are recognised by industry players.
Erich Andersen, corporate vice president and chief IP counsel at Microsoft, said that the increasingly nationalistic and populist environment, as shown by US-China trade tensions and Brexit, is worrying. He said that there needs to be a strengthening of multilateral agreements and treaties that go beyond national boundaries.
“Nationalistic movements are a source of great frustration for companies that have invested but are unable to recoup investments,” said Andersen. But he remains optimistic as changes in IP systems, such as those in China, are helping to close the gap between IP regulations across jurisdictions.
Data-sharing and AI
Another source of concern for Andersen is that a large part of artificial intelligence (AI) is not patentable because it is fundamentally mathematical formulae. He said that trade secrets are often a better form of protection, but they must be actual secrets, and businesses need the technical security to protect them.
He said that data – the fuel of AI – can be protected by copyright. However, a challenge remains in training AI for fair use, such as whether algorithms may use copyright-protected works without permission from authors. Another challenge is that data-sharing systems remain too cumbersome. Although open source software is available, data-sharing standards are not.
“The world needs more standards and norms for sharing data,” said Andersen. “Researchers feel blocked from doing research that will have value for society. Even though laws are there for privacy, those to encourage open use are not there, and this introduces risks to projects.”
He added: “In order to make our systems more valuable, we need to share data, whether it is in medical research or fighting against climate change. Our society needs large data sets to figure out these solutions.”
According to Lyse Brillouet, vice president of IP and licensing at Orange Group in France, the digital environment will be increasingly challenging as devices become more connected, but incubation periods for products are still long in a rapidly changing environment. While the company has more than 7,000 patents spanning wireless, mobile services, IoT, AI, big data, security, infrastructure and more, it has been focused traditionally on protecting patents for defensive purposes.
This is gradually changing, Brillouet said, explaining that while IP is reserved first and foremost for inventions, it has to be focused on value to create partnerships. IP departments of the future also need to collaborate with more business units such as marketing and engineering for them to better understand IP value and strategy. However, existing paradigms pose challenges to open innovation and there are a lack of soft skills such as communication and education, she concluded.