Data from Patent Strategy’s Machine yearning: AI and patents survey, which polled more than 60 in-house lawyers across different industry sectors, revealed that most businesses (59%) have either started to collaborate with an AI developer to create a new product or are planning to do so in the next five years. Only 10% of respondents said that they would never enter into such a collaboration.
Sources explain that non-tech businesses have gone into, or are planning to go into, AI-developer partnerships because they want to use the technology to improve upon or create new products but do not have the necessary in-house expertise.
But those collaborations have raised IP-related challenges for patent-focused businesses, and particularly for those businesses that are not used to partnering with firms outside their own industry.
Their biggest IP-related concern in such a partnership, according to the Machine yearning poll, was IP ownership. Out of the 28 in-house lawyers who answered the question on what their biggest challenge was or would be in an AI collaboration, 16 respondents (57%) cited IP ownership.
“The biggest challenge would largely depend on the agreement between the developer and company and would boil down to the negotiation of IP ownership,” says the IP manager at an automotive start-up.
The most pressing challenge after that, with almost a third (32%) of the vote from survey respondents, was developing an effective IP strategy between the two parties.
“The specifics of the situation would dominate over any AI-related issues,” says the IP manager at an electronics company. “As with any technology, issues of patentability, inventorship, associated un-registerable IP, such as know-how and trade secrets, rights to own, rights to commercialise, and such, will throw up a different pattern of challenges depending on the people and company cultures involved.”
Secret strategiesSources say that while ownership is a big concern for most companies when it comes to AI collaborations, that challenge is not specific to the technology.
“As long as the contract addresses IP properly, I do not see this as any more challenging than a traditional collaboration,” says the head of IP at a government department. “A clear agreement on inventorship and ownership of patentable inventions will be important in any partnership.”
Counsel have previously told Patent Strategy that such agreements should cover who will own the IP, whether it will be jointly owned and whether each party has the same rights to commercially exploit the invention without paying money to the other side.
What businesses are perhaps more concerned about when it comes specifically to AI collaborations is how to work together to ensure their joint innovations are adequately protected.
The IP adviser at a global automotive manufacturer points out that AI is difficult to patent. He adds that even if companies manage to patent the technology, enforcement is also difficult because infringement is so hard to detect in these ‘black box’ or ‘invisible’ technologies.
“If you were to obtain patent protection, how could you know whether a third party had infringed on that invention?” adds the patent attorney at another global automotive company.
Non-tech companies and AI developers thus need to work together to decide which parts of their joint IP should be formally protected by patents and which should be held as trade secrets. This can lead to strategy squabbles.
The automotive IP adviser says that larger companies may be happy to keep AI-related inventions as trade secrets if they already have large and impressive portfolios. That may not be the case for AI developers, however, if they are start-ups or SMEs that are looking to build up their patent portfolios.
“Start-ups like having patents,” he says. “They see patents as more than enforceable assets. They see them as something to be proud of that can be leveraged as a marketable asset for the company.” He adds that the question then becomes how those businesses could replace that asset if it was decided that a joint venture would be protected by trade secrets.
One source points out, however, that sharing trade secrets and confidential matters between two different companies is tricky. Confidentiality and trade secrets law in most jurisdictions, such as the EU (which has its Trade Secrets Regulations), requires that the subject matter be seen by the business as a secret and that it is treated as a secret.
One party in the partnership may have the right processes and procedures in place to stop these secrets getting into the wrong hands and to give them a legal avenue to follow should those preventions ultimately fail. Those procedures won’t prevent another company stealing and commercialising the company’s lucrative AI-enabled innovations, however, if the other business does not have the equivalent measures implemented.
SME for me?Patent Strategy’s Machine yearning report also showed that 5% of respondents were creating a new AI-enabled product but were doing so without the help of an AI developer. Sources suggest that this situation could become increasingly common as businesses acquire AI companies to bring the necessary expertise in house.
The head of IP in the medical device arm of a life sciences business points out that the complexities of determining IP ownership and protection strategies between two entities could lead some companies to build their own AI-savvy resources.
The automotive IP adviser says that there is no history of swallowing up tech companies in the automotive industry, but that M&A is a common way of acquiring skills and assets and that car manufacturers may have to change their view on the matter.
The patent counsel at another automotive company says that M&A will likely become more common as car companies look to bring the necessary tech expertise in house, and they will increasingly have to be vigilant when it comes to choosing the right partner and conducting effective due diligence.
More businesses are banding together with tech developers as they seek to leverage the potential of AI technologies. But that entanglement comes at a price – particularly when it comes to working out what to protect and how to protect it.
The material on this site is for law firms, companies and other IP specialists. It is for information only. Please read our Terms and Conditions and Privacy Notice before using the site. All material subject to strictly enforced copyright laws.
© 2020 Euromoney Institutional Investor PLC. For help please see our FAQs.