The road ahead – IP challenges in the automotive industry
Anthony Del Monaco, Kara Specht, and Kathryn Judson of Finnegan weigh in on AI-related technologies and other IP challenges facing car companies
Since the automobile became a primary mode of transportation, emerging technologies and ensuing consumer demands have redefined the way vehicles are manufactured and operated.
At the beginning of the 21st century, consumers mostly based their decisions to purchase cars on factors such as engine specifications, vehicle reliability, and sticker price. As time went on, they started to consider other factors, including advanced safety features and augmented reality capabilities.
Now, consumers want a customised and personalised experience and to feel connected to their vehicles. Automotive companies have endeavoured to address this shift by focusing on the customer experience, and by integrating digitalisation and automation into their manufacturing processes and product development.
Many vehicle manufacturers now support technology that allows cars to park themselves. Other car firms have created gesture control systems that detect hand movements, allowing drivers to change the volume of playing audio with a simple circular motion.
Car companies are also innovating in digital technologies, including big data and predictive maintenance, to improve manufacturing efficiency and consumer safety.
But as the automotive industry has embraced more and more innovation, it’s had to deal with more intellectual property challenges and considerations.
Recent federal court decisions and artificial intelligence-related patent rejections at the USPTO, for example, have left automotive innovators with uncertainty over whether US IP law will protect their innovations.
The car industry also has to consider the possibility that it will see an increase in patent challenges as it continues to embrace AI.
Patent eligibility challenges
One of the biggest challenges automotive innovators face at the moment is working out whether their claimed inventions qualify as patent-eligible subject matter.
In the US, the USPTO frequently strikes down AI-related innovations under Section 101 in Title 35 of the US Code for lack of patent eligibility.
Section 101 sets out that “any new and useful process, machine, manufacture, or composition of matter” is eligible for patent protection.
But the Supreme Court (SCOTUS) recognised several judicial exceptions under the statute, including claims directed to abstract ideas, laws of nature, or natural phenomenon.
Under SCOTUS’s two-part test from Alice Corp v CLS Bank in 2014 and Mayo Collaborative Services v Prometheus Laboratories in 2012, a claim is ineligible under Section 101 where it is “directed to” a law of nature, natural phenomenon, or abstract idea, and the “claim’s elements, considered both individually and ‘as an ordered combination,’” are insufficient to “‘transform the nature of the claim’ into a patent-eligible application”.
Since these decisions, those seeking patent protection in the US have navigated through ruling after ruling to interpret whether ideas are patent eligible or ineligible.
The USPTO and federal district courts have used Section 101 to invalidate numerous patents on digital technology and AI since Alice.
A divided panel at the Court of Appeals for the Federal Circuit denied rehearing on American Axle & Manufacturing v Neapco Holdings, a case involving driveshaft automotive technology, on July 31, 2020.
In this matter, the Federal Circuit held that certain claims relating to a method for manufacturing were patent ineligible under 101 because the claims were directed to a law of nature.
Notably, Judge Pauline Newman stated in her dissent that the “court’s rulings on patent eligibility [had] become so diverse and unpredictable as to have a serious effect on the innovation incentive in all fields of technology”.
The Federal Circuit’s American Axle decision brings increased attention to the patent eligibility problem in the automotive industry, beyond those sectors traditionally subject to such problems including biotech and software.
In previous years, the automotive industry faced fewer patent-eligibility challenges because innovations were typically directed to physical objects or processes.
But with new innovations involving machine learning, autonomous driving, and other software-based AI processes coming from it all the time now, the automotive industry will face more patent-eligibility challenges in the future.
If the Supreme Court grants American Axle’s pending writ of certiorari, this may provide additional guidance to patent applicants in automotive with respect to Section 101.
AI inventorship issues
Then there’s the matter of the rules surrounding AI inventorship and patentability in the US.
AI-related inventions in the automotive industry span a wide range of functions, including autonomous navigation, advanced driver-assistance system, battery management, and speech recognition.
But with AI providing more and more information and solutions, the original ‘inventor’ is becoming further removed from the final invention.
Around the world, patent offices have reached different conclusions about whether an AI system can be classified as an inventor or not.
For example, Stephen Thaler filed two patent applications listing DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) as an inventor in patent offices around the world, including those in the US, Europe, Australia and South Africa. The AI system devised the invention with no human input.
In April 2020, however, the USPTO concluded that only a natural person (and not an AI system) could be named as an inventor on a patent.
In rejecting the applications, the USPTO found that because current statutes, case law, and USPTO regulations limited inventorship to “natural persons”, the applications failed to properly identify an inventor under Section 115a in Title 35 of the US Code.
Until recently, every country that had considered the DABUS applications rejected them, including the USPTO, the UKIPO and the EPO. Then in July 2021, the South African patent office surprised the global community when it granted a patent to the same DABUS system.
The Federal Court of Australia then came to a similar decision that AI was eligible to be designated as a patent inventor.
We have yet to see the impact of these decisions, but they complicate predictability for car companies in jurisdictions outside of the US looking to expand protection in the country, where AI cannot be an inventor.
Although Thaler experienced some success in South Africa and Australia, his appeal of the USPTO’s rejection of his patent applications in a US district court failed.
In Thaler v Hirshfeld, the court found that the “issue of whether an artificial intelligence machine can be an ‘inventor’ turns on the plain meaning of the statutory term ‘individual’”. Thus, the court affirmed the USPTO’s determination that an inventor must be an individual.
Thaler indicated that he planned to appeal the decision to the Federal Circuit. But for now, the ruling highlights the legal and legislative debate and uncertainty surrounding AI technologies and inventorship.
Therefore, innovators may still have many lingering questions before filing for patent protection.
For example, could a patent be subject to an inventorship challenge in cases where an automotive AI system generated only a portion of an invention autonomously?
Should the inventor of the underlying AI technology be named as an inventor downstream of an AI system’s innovations?
Trade secret protection
When considering ways to protect digital and AI-related innovations in the automotive industry, companies typically turn to patent protection.
But trade secrets also provide a valuable form of protection, and a possible counterbalance to AI-inventorship concerns.
Trade secrets can, and usually do, supplement a patent portfolio. They can also independently provide protection for innovations, especially when a patent may be difficult to obtain because of patent-eligibility and inventorship concerns.
Unlike patents, trade secrets can be retained indefinitely so long as the confidential and proprietary information remains secret and the owner uses reasonable measures to maintain secrecy.
While an appealing option, trade secret protection comes with its set of potential drawbacks.
For one thing, trade secrets are only enforceable if the secret is properly protected and misappropriated by another. For another, trade secrets cannot be enforced against others who discovered the technology through reverse engineering or who independently created an identical invention. Accidental disclosure also destroys a trade secret.
Automotive companies can encounter difficulties in both properly maintaining secrecy and in proving misappropriation. These difficulties often arise because of the volume of proprietary information a typical car company possesses, the mobility of the workforce, and the collaborative nature companies have with potential competitors or joint venturers.
To ensure they adequately protect their trade secrets in AI and digital technologies, companies must first understand the full scope of what constitutes a trade secret.
This scope could include manufacturing techniques, algorithms, and any other commercially valuable confidential and proprietary information. If a company fails to appreciate the breadth of what a trade secret could protect, it might not protect the full range of the trade secret.
To no surprise, most trade secret misappropriation involves a company’s employees or business partners leaving to work for a competitor or starting their own competitive firm.
For example, in January 2021, Tesla filed a motion for a temporary restraining order and a complaint for trade secret misappropriation and other claims against a software automation engineer who had been with the company for only three days.
In Tesla v Khatilov, filed at the District Court for the Northern District of California, Tesla alleged that a newly hired software engineer stole thousands of files of code and began downloading proprietary information from its networks onto personal devices through Dropbox.
Tesla’s information security team detected the download through its monitoring software and had a non‑disclosure agreement in place prohibiting the use or disclosure of proprietary information, including technical data.
The Tesla case highlights the importance of protecting trade secrets by implementing reasonable measures to maintain secrecy of a company’s proprietary information.
Companies may take different measures to keep the information protected, such as limiting access to trade secrets to only employees who absolutely need to know them and making sure only a select few individuals know the entire scope of the trade secret.
They could also secure physical facilities, limit access by employing passwords and encryption only for authorised users, limit access of company files from non-secure or public networks, require employees to return laptops and files when they leave the company, and prevent employees from copying or deleting any data or files.
By proactively addressing the intellectual property concerns discussed, car companies should leverage the full scope of available IP protection to maintain ownership and commercialise their investment in digitalisation and AI innovations.
As AI takes a key role in the manufacture and design of inventions and systems integrated into automobiles, industry leaders need to re-think creative processes and be ready to respond to the challenges yet to come.
This article was authored by Anthony Del Monaco (a partner in Washington DC) Kara Specht (a partner in Atlanta) and Kathryn Judson (an associate in Atlanta) at Finnegan