Artificial intelligence (AI) is poised to transform every sector of the economy, from healthcare inventions like Autism Virtual Reality, which lets people experience the sensory world of an autistic child, to agricultural drone imagery that alerts farmers about poorly irrigated farmland. The global AI market is projected to reach $1.6 billion by 2030, making it one of the fastest-growing sectors and one that holds the most promise for businesses looking to protect their inventions.
When considering new technological frontiers, it’s always interesting to dig into what the largest tech giants in the world are investing in and protecting. Meta launched Make-A-Video, an AI program that turns text prompts into high-quality videos.
Similarly, generative AI has launched on TikTok to help creators make virtual reality filters. Microsoft is heavily invested in AI research, building integrated systems and embracing reinforcement learning, which allows human researchers to interact with AI models to help them learn. Apple has recently acquired two AI tech startups; one focused on image recognition and the other on voice interaction. Even amidst significant job cuts at Alphabet, the tech giant is doubling down on AI efforts.
The Growth of Artificial Intelligence
Investments in AI startups have grown 6X since 2000, and global GDP is expected to increase to $15.7 trillion by 2030 in part due to AI. As of 2022, the global AI market sits at a value of $136.6 billion, and countries around the world are racing to take their place at the top of the AI leaderboard.
77% of companies are using or exploring AI, and large enterprise organizations are twice as likely to have adopted AI. Specific industries like financial services, healthcare, and automotive have sped ahead when it comes to AI adoption, but manufacturing, consumer goods, and retail sectors are quickly catching up.
AI adoption promises significant benefits across every single industry; from a high-level, increased efficiency, productivity, and accessibility make huge impacts on a business’s bottom line. AI can reduce costs, drive optimizations when it comes to resources, and provide a level of interconnectivity to dispersed systems and programs.
As AI technology and intellectual property stand to be one of a business’s most valuable assets moving forward, there is increased competition around protecting these systems. How can companies ensure their inventions are properly safeguarded against competitors? How should organizations leverage these innovations to drive even more growth and revenue? For many businesses, the answer usually boils down to “patents.” More than 100,000 organizations own patents, and for some of the largest tech organizations like Samsung and IBM, patents are incredibly valuable intellectual property assets.
Can Integration of Artificial Intelligence Technology Within Your Technology Be Patented?
First, it’s important to understand what makes AI intellectual property (IP). The concept of “IP” rewards and incentives innovation and creation, and AI is no different. IP is defined as a creation of the mind, from literary works to logos and symbols. Understandably, just like how companies patent so many other proprietary inventions to protect them from illegal usage, AI creators want to safeguard their intellectual property.
Several court cases have concluded that “No, AI is not patentable.” In Thaler v. Vidal, the US Supreme Court ruled in September 2021 that inventors were limited to humans. The US Patent Act also defined an “inventor” as one “who invented or discovered the subject matter of the invention.” Similarly, the UK, European, Australian, and German patent courts have crafted similar rulings, naming humans as the only “inventors.”
However, this certainly hasn’t stopped companies from racing to safeguard their IP in the AI patent boom. Businesses want to protect their inventions against competitors while also building their own differentiated edge over others. The research in “Inventing AI: Tracing the diffusion of artificial intelligence with U.S. patents” from the United States Trademark and Patent Office (USTPO) shows that the widespread adoption of new technology brings an inevitable wave of patents, and that’s exactly what’s happening in the world of AI.
Patent filing has increased by 100%+ in the last 16 years and now makes up a whopping 42% of technology filing subclasses. The highest growth rate of patent filings happened between Q1 and Q2 of 2022, showing a record number of AI patents hitting the market. While it’s not guaranteed any of these will be granted, the race is on to see who and what can be successfully claimed.
Benefits of Obtaining Artificial Intelligence Patents
So, what are the benefits of obtaining artificial intelligence patents? Any sort of legal filing process takes time, resources, and money, so it’s critical for businesses to gain actual value from filing. First, many companies simply file to validate the uniqueness of their invention. Companies can use “patent-pending” in marketing language, making the technology seem valuable and highly proprietary.
Patents can signal business growth and innovation, which is important to investors and strategic partners. Additionally, once a patent is actually successfully claimed, it protects this “secret sauce” from illegal use from others and helps a company stand out from the competition.
However, filing a patent does come with some risks. First, patents are visible while they are being filed, exposing “secrets” to a wide audience. Also, without the right legal counsel, many patents are often rejected, leaving companies wasting significant time and money.
Identifying Patentable AI Inventions
If you’ve decided the risks outweigh the reward and you’re ready to secure an AI patent, you need to determine if it’s patentable. To determine if your invention is patentable, examine the IP landscape to ensure your claim isn’t infringing on another. Some AI can be identified by building block functions.
Functional claiming comes into play when an inventor can claim what a program does versus reciting its structure. This allows AI inventors to claim ownership over a variety of specific solutions by patenting the core functionality.
Additionally, there is a list of requirements to determine if an invention is patentable. It must:
• Be made by human inventors
• Be a novel idea, meaning “new”
• Be an inventive step or “non-obvious” — the invention must show creativity and would not be common knowledge to someone in that field
• Offer industrial application — the invention must have economic significance or be able to be used in some manner
Understanding Patentability Requirements
Once you’ve determined your AI is patentable, it’s time to understand patentability requirements. Like any patent, it must be described in a way that deems the technology an “invention.” In Alice v. CLS Bank, the US courts decided a “patent-eligible” claim could not be “directed to an abstract idea” and, instead, needs to include an “inventive concept.”
In 2016, the Supreme Court ruled in Enfish v. Microsoft that “[s]oftware can make non-abstract improvements to computer technology just as hardware improvements can,” giving hope to those seeking patents for AI. Additionally, arguing for one highly detailed technical claim will increase your chances of winning your case.
When determining how to write your patent application, keep in mind the different types of infringement you’re trying to protect against. The two types of infringement are direct and indirect. As you would expect, direct infringement showcases intent, whereas indirect can occur without prior planning or knowledge of the infringed-upon invention.
Applying for Patent Protection
There are a few different avenues when applying for patent protection. In Alice Corp. Pty. Ltd. v. CLS Bank International, the ruling created a two-part test for determining patent eligibility. First, it asks whether claims are directed to a patent-ineligible abstract idea, natural phenomenon, or law of nature. Second, the claim must contain an “inventive concept.” The ruling on “abstract ideas” posed a challenge for many AI and ML technologies as they often rely on algorithms or mathematical formulas to work.
Luckily, some recent cases have overruled these precedents and offered hope to AI inventors. CardioNet, LLC v. InfoBionic, Inc. actually reversed a decision that dismissed a previous claim of patent infringement on an AI cardiac technology. The new ruling determined that the AI innovation improved cardiac monitoring devices and was, therefore, an inventive concept.
These cases illustrate the importance of crafting a specific claim that demonstrates a technological improvement with more than just basic data manipulation; these successful claims further build on how these new concepts improve on previous technologies. These overrulings and dismissals show why it’s critical to work with an experienced AI patent attorney from the beginning.
Start With an AI Patent Search
First, start with an AI patent search to make sure you’re not trying to overwrite an existing claim. A search not only prevents infringement but also provides material and precedents for building a strong case. Get a better understanding of where AI technology fits into the landscape and how similar inventions have fared in the application process. If the AI is still in development, a patent search is a good place for research and development and helps gather analytical data of competitors or similar technologies.
Google offers a free patent search function through Google Patents, and The Patent Public Search tool is hosted for free as well by the U.S. government. However, while these tools seem simple, deep research requires hours of work from an expert attorney who is well-versed in exactly what to look for. The first filing of a patent is critical as all software information must be disclosed at this time; adding new information at a later date can jeopardize the application, thus working with an experienced AI IP lawyer is key.
Developing an AI Patent Portfolio
Building a strategic AI portfolio is complex and requires intense forethought. Most businesses at least know the initial software or system that is patentable, but how do you keep building continuously on these innovations? Especially when resources are limited, filing multiple patents in a “throwing spaghetti at a wall to see what sticks” strategy is not effective. Determine the order of importance of your inventions.
A prior art search shows if your features or software have already been filed. Often, this saves a ton of time and money as quick searching shows that some systems have already been patented. However, a prior art search will show you the breadth and depth of other successful claims, giving you inspiration and knowledge to further yours.
When thinking long-term, there are two ways to build an AI portfolio: offense and defense. A defensive patent strategy looks to prevent competitors from maintaining patents. On the other hand, when thinking about AI patents, most people imagine an offensive patent strategy. This is when companies file for patents in an effort to license their technology at a later date. Also, this gives companies the power to bring a potential lawsuit against an infringer to either stop their sales or force the other company to pay them royalties.
Artificial Intelligence Intellectual Property Issues
While it’s exciting to think about an offensive or defensive patent strategy, the process is absolutely not without challenges. First, many people argue for a human-centered approach to IP; humans are the ultimate creators and inventors, but what happens if AI invents its own intellectual property?
58% of people believe the company that owns the AI should hold the rights, while only 4% would credit the AI itself. Also, as AI becomes more sophisticated, systems are actually creating what would be considered art; for example, Magenta is an open-source research project that is exploring the role of machine learning in art creation — and actually creating art in the process!
There are so many intricacies and unexplored areas when it comes to AI patents. As new technologies emerge every year, patent application and intellectual property issues will only get trickier.
Can an AI System Itself Infringe a Patent Claim?
There are two scenarios where an AI system itself could infringe on a patent claim. The first is if it is performing the identical steps of another system. This is a direct infringement and can be prosecuted as such. The entity responsible for that performance would be held liable; more often than not, this is the company that owns the AI in question.
The second scenario where an AI system itself could infringe on a patent claim is through indirect infringement; however, this once again becomes a sticky situation in determining if the AI had prior knowledge and intent, which is almost impossible to determine.
How To Secure An Artificial Intelligence Patent And Safeguard Your Innovations
New innovations across every sector are popping up every month, and successfully filing an AI patent is a time-consuming, complex process with lots of nuances. From determining if an invention is even eligible to digging through hundreds of thousands of previously filed patents to making a strong, successful case, without the right support filing, it’s difficult to feel the peace of mind and confidence that your invention is fully protected to its full extent.
Working with an expert team of experienced patent attorneys, specifically in AI, is the right decision when it comes to protecting one of your most valuable assets. At the Rapacke Law Group we will walk you through the type of IP protection that might be the best for your situation and explain every step of the process.
To learn more about protecting your AI inventions, book a strategy call with one of our expert attorneys today.