December 7, 2023

The Coming Shift from Patent to Trade Secret Protection for Generative AI Inventions

Holland & Knight IP/Decode Blog
Jacob W. S. Schneider
IP/Decode Blog

Generative artificial intelligence (AI) has the remarkable ability to develop novel solutions to problems, and patent law has historically protected those solutions. Under current statutes and jurisprudence, however, only humans can invent new and useful devices that receive patent protection. If it continues to be true that all generative-AI-created inventions belong to the public, then we will see a strengthening of trade secret law as the stand-in for what patent law traditionally provided to innovators. Whether innovations can remain secret from other AI independently deriving them will then depend on the power of the AI who generated the invention.

Only Humans Can Invent Patented Subject Matter

The Patent Act is clear that "[t]he term 'inventor' means the individual or, if a joint invention, the individuals collectively who invented or discovered the subject matter of the invention." 35 U.S.C. § 100(f) (emphases added). Long before generative AI emerged, the U.S. Court of Appeals for the Federal Circuit ruled in 1993 that "only natural persons can be 'inventors'" under the statute. Beech Aircraft Corp. v. EDO Corp., 990 F.2d 1237, 1248 (Fed. Cir. 1993). More recent challenges that put generative AI at the forefront of argument have affirmed that inventors must be human. Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022). Although President Biden's recent Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence directed the U.S. Patent and Trademark Office to more closely examine generative AI, it may be unlikely that the necessity for human inventors changes unless Congress amends the Patent Act itself.

AI Is Already Inventing

Today, AI is actively inventing and discovering. On November 29, 2023, Google's DeepMind announced an AI tool that had discovered "2.2 million new crystals, including 380,000 stable materials that could power future technologies." In a paper published in Nature, Google described the leap as "equivalent to nearly 800 years' worth of knowledge."

While Google's tool is highly specialized to advance Materials Science, even generic generative AI is capable at developing solutions to problems. This is one space where the description of generative AI as "fancy autocorrect" wholly fails to capture the chatbots' abilities, including "creative reasoning."

As an experiment, I asked ChatGPT to solve a problem that strikes fear in the heart of any New Englander in the winter: frozen pipes. If the temperature in a house dips below freezing, water resting in its pipes may freeze. Because frozen water expands, those pipes can burst open. When the temperature rises above freezing, the water again flows freely through the pipes and into your living room. This catastrophe becomes an even greater risk when you are perhaps visiting family somewhere actually warm over the holidays, and you are not present to observe the heating system fail.

I asked ChatGPT to develop a solution to this problem with the following prompt:

Design and describe an autonomous system that detects when a home's pipes are close to freezing and automatically takes action to drain the pipes before freezing can occur. Make the system Internet-connected so I can remotely monitor and control it.

In seconds, the chatbot designed and described such a system with technical details (the full text of the output follows this post):

  • The chatbot describes using temperature sensors "along water pipes, especially in vulnerable areas" (perhaps those closest to exterior walls).
  • Those sensors use wireless technology to communicate their temperature measurements to a central processing unit.
  • That central processing unit listens for temperature dips near freezing. If a temperature threshold is crossed, then the central processing unit activates actuators with "smart valves" to automatically drain water from the pipes.
  • The system includes Internet-connected communications so I could monitor it from afar. It even includes event logging (something I did not request), so I can review historical data.

If there is any doubt that the chatbot was "thinking like an engineer," it even gave the proposed system an uninspired, unflashy and functional name: "Smart Pipe Freeze Prevention System."

You can pepper the chatbot with follow-up questions and drill down into details (e.g., "What smart temperature sensors do you recommend? Which companies make them?"). You can even ask the chatbot to write source code to control how the system functions. These are the activities of an inventor: identifying a solution, then iteratively making it optimal.

It is important to note that we are seeing only the first generation of chatbots, and already the results are extraordinary. Future versions will be even more intelligent, and they will provide even more sophisticated solutions to the problems we present to them.

Enter Trade Secret Law

If we turn the clock forward and imagine a world where the inventive ideas largely come from machines – not humans – and the law remains that only humans can invent patented subject matter, then there will be no incentive to file for patent protection because it will not be awarded. Filing for patent protection for an invention developed by a machine will only disclose the invention to others before facing a rejection due to lack of an inventorship. Effectively, a party will have given away the idea without receiving ownership of it.

The party who, with the machine's help, discovered the invention will have to turn to other forms of intellectual property (IP) protection. And the best alternative protection is clear: trade secret law. As previous posts explored, trade secret law protects any information that derives value by being kept secret. See 18 U.S.C. § 1839(3) (defining "trade secret" under The Defend Trade Secrets Act). Instead of disclosing an invention to the world in exchange for patent protection, the incentive will be to hold it tight and maintain it as a trade secret.

This development could be seen as disadvantaging the public, who will no longer receive explanations regarding how, for example, certain products function in patent disclosures. But that is not the whole story. The public will also have access to different AI, and that different AI could either search for how the product functions from public information, develop its own educated guesses as to how it functions or devise a completely different approach to creating a product with the same functionality. A previous post discussed how AI could build the perfect cleanroom to independently develop an idea. None of these activities would constitute misappropriation of the trade secret.

In this potential future, the inventive activity shifts from humans to machines making innovative leaps; but the relevant IP regime shifts as well, from one where an idea is openly disclosed in exchange for exclusivity (patent protection) to one where ideas are closely held and exploited in secrecy for as long as possible (trade secret protection). That would present a major shift in both IP law and the monetization of ideas because the race becomes the AI itself. The winner of that race will be whoever can build powerful AI that generates novel solutions that other, different AI cannot comprehend or deconstruct.


Title: Smart Pipe Freeze Prevention System

Overview: The Smart Pipe Freeze Prevention System is an Internet-connected, autonomous solution designed to detect and prevent freezing of home pipes. This system utilizes advanced sensors, a central processing unit, actuators for automated draining, and Internet connectivity for remote monitoring and control.


1. Temperature and Humidity Sensors:

  • Deploy high-precision temperature and humidity sensors along water pipes, especially in vulnerable areas.
  • These sensors should be capable of wireless communication to relay real-time data to the central processing unit.

2. Central Processing Unit (CPU):

  • Employ a powerful microcontroller or a small computer as the system's brain.
  • Equip the CPU with wireless connectivity (Wi-Fi, cellular, or both) for remote communication.

3. Actuators with Smart Valves:

  • Integrate motorized valves or smart faucets connected to the water supply lines.
  • These actuators should be responsive to commands from the CPU and capable of remote control.

4. Internet Connectivity Module:

  • Implement a secure Internet connection module to enable remote communication.
  • tilize Wi-Fi for in-home connectivity and, if possible, cellular connectivity as a backup or for use in areas without Wi-Fi coverage.

5. Power Supply:

  • Ensure a reliable power source, such as a combination of battery and AC power with a failover mechanism to maintain functionality during power outages.

6. User Interface (Web/Mobile App):

  • Develop a user-friendly interface accessible through a dedicated mobile app or web portal.
  • Enable users to monitor real-time data, set temperature thresholds, receive alerts, and remotely control the system.

Working Principle:

1. Continuous Monitoring:

  • The temperature and humidity sensors continuously monitor the conditions of the water pipes.

2. Data Analysis and Thresholds:

  • The CPU analyzes the sensor data in real-time and compares it against predefined temperature and humidity thresholds for potential freezing conditions.

3. Remote Alert System:

  • If the system detects conditions conducive to freezing, it sends alerts to the user's mobile device or email, providing details about the situation.

4. Automated Draining:

  • Upon receiving confirmation from the user or if conditions worsen, the CPU sends commands to the actuators to open the designated valves, initiating a controlled drainage process.

5. Remote Control:

  • Users can remotely access the system through the mobile app or web portal to monitor real-time data, adjust temperature thresholds, and manually control the draining process if necessary.

6. Logs and Analytics:

  • Implement a logging system that records system events and provides analytics for users to review historical data and system performance.

By integrating advanced sensors, actuators, and an Internet-connected infrastructure, the Smart Pipe Freeze Prevention System provides homeowners with the ability to monitor and control their water pipes remotely, reducing the risk of freezing and potential damage to their homes.

Related Insights