Patent Issues Involving AI-Engineered Solar-Foam Cooling Pavements.

01 Apr 2026 --
0 Comments

📌 1. Inventorship and AI — DABUS Litigation (Thaler Cases)

Background

The Device for the Autonomous Bootstrapping of Unified Sentience (DABUS) is an advanced AI system developed by Dr. Stephen Thaler to autonomously generate inventions. Beginning in 2019, Thaler submitted patent applications listing DABUS as the sole inventor in jurisdictions around the world. These cases have become the foundational test of how patent systems treat AI‑generated inventions.

Case: Thaler v. USPTO / Thaler v. Vidal (United States, Federal Circuit)

Core Legal Issue: Can an AI system be named as an inventor under U.S. patent law?

Facts: Thaler filed U.S. patent applications claiming inventions conceived by DABUS and listed the AI system as the inventor — without any named natural person (i.e., human) as the inventor.

Holding: The U.S. Court of Appeals for the Federal Circuit ruled that only natural persons (humans) can be named inventors under current U.S. law. AI systems, even if they generate inventions autonomously, are not recognized as inventors. This follows the statutory requirement that inventorship requires a conception — a mental act that only a human can perform. The U.S. Supreme Court later declined to hear any appeal, leaving the Federal Circuit decision intact.

Implication for Solar‑Foam Cooling Pavement:
If an AI engine independently generated a novel foam cooling pavement technology, a patent application cannot be filed naming the AI as the inventor in the U.S. Instead, a human will have to be shown to have conceived the invention to obtain a valid patent.

Case: Thaler & DABUS Patents (South Africa & Australia)

South Africa Patent Office (2021):
South Africa became the first jurisdiction to grant a patent on inventions listing an AI as inventor — with the AI’s owner as the patent owner. This was significant because it diverged from more conservative jurisdictions like the U.S. and Europe.

Federal Court of Australia (2021):
Similarly, an Australian Federal Court ruled that AI‑generated inventions could be “inventive” and patentable, allowing DABUS to be listed as an inventor under Australian law. This showed that some jurisdictions, at least historically, permitted expanded notions of inventorship.

Implication:
Patent eligibility and inventorship rules are jurisdiction‑specific. An AI‑engineered solar‑foam cooling asphalt might be patentable in some countries but not in others, based on how local patent law treats AI inventorship.

📌 2. Inventorship Standards and Human Contribution: USPTO Guidance

USPTO 2025 Inventorship Guidance

The U.S. Patent and Trademark Office (USPTO) issued updated guidance confirming:

AI cannot be listed as an inventor.
Human inventors must demonstrate that they conceived the core inventive concept.
AI systems are treated like tools (e.g., software or lab equipment).

Inventorship Test in Practice:
To obtain a valid patent for an AI‑assisted invention (like an AI‑designed cooling pavement compound), the applicant must show:

Conception — the human had a definite and permanent idea of the complete invention before AI refinement.
Reduction to Practice — the human was actively involved in developing or enabling the invention beyond mere AI operation.

Legal Issue Example:
If an engineer merely “pressed a button” and the AI produced the final chemical formulation for solar‑foam cooling, that might be legally insufficient human contribution. The inventor must actively guide the AI and provide inventive insight.

📌 3. Patentability of AI‑Related Technology Under §101 — Alice Framework

Even if inventorship is solved, many AI inventions face a separate eligibility hurdle under patent law — particularly in the U.S.

Case: Recentive Analytics, Inc. v. Fox Corp. (Federal Circuit, 2025)

Issue: Whether patents involving machine learning for content scheduling and placement were directed to an “abstract idea” and thus unpatentable under Section 101 of U.S. law (which prohibits patents on abstract ideas without an “inventive concept”).

Holding: The Federal Circuit reaffirmed that simply applying machine learning in a new domain is not enough. The claimed innovation must show a specific technical improvement or inventive concept beyond the abstract idea itself.

Implication for Solar‑Foam Cooling Pavements:
Patent claims should avoid abstract descriptions (like “AI optimizes cooling”) and instead include concrete, technical steps — e.g., a specific foam composition and process steps that show inventive technical contribution beyond general AI modeling.

📌 4. European Patent Office and International Practice

DABUS at the European Patent Office (EPO)

Issue: Like the U.S., the EPO rejected patent applications listing a non‑human inventor. The EPO insisted that inventorship must be a natural person for titles of patent rights. This aligns with established European patent practices focusing on human inventors.

Implication:
Patent applications on AI‑assisted solar‑foam pavement technologies at the EPO must similarly identify human inventorship.

📌 5. UK Supreme Court: Patent Eligibility of AI Systems (Indirectly Relevant)

UK Supreme Court Ruling (2026)

Issue: Whether computer programs, including neural networks, can be patented.
Holding: The UK Supreme Court ruled that AI systems involving underlying hardware can, in principle, be patentable — overturning the outdated notion that software alone is unpatentable.

Relevance:
This broadens patent eligibility for AI architectural innovations, which could indirectly benefit inventions like an AI engine that designs cooling pavements, so long as the patent claims tie the AI improvement to physical, technical hardware effects — for example, how the AI’s output affects a cooling pavement’s structure or performance.

📌 6. Patent Ownership and Assignment Disputes (General Principles + Pannu Case Reference)

While not always high‑profile cases, disputes over who owns the patent rights can occur even when human inventorship is clear.

Pannu v. Iolab Corp. (Inventorship Correction Case)

Issue: In U.S. courts, the Pannu case addressed whether a named inventor left off a patent should be added because they contributed inventive subject matter. While this case predates AI, it illustrates how inventorship disputes are resolved.

Principle: If someone significantly contributed to conception, they must be named. This is relevant if an AI‑assisted project involved a human who was wrongly omitted.

Implication for AI Use:
Multiple humans may be involved in designing, training, and applying AI to a cooling compound. Ownership disputes can arise, and courts will examine who actually contributed to the inventive concept.

📌 7. Key Themes & Lessons for AI‑Engineered Solar‑Foam Cooling Pavements

A. Inventorship Is Human‑Centric

Globally, most patent systems require human inventors. AI is a tool, not a listed inventor. You must document human insight into the invention.

B. Patent Eligibility Must Overcome Abstract Idea Barriers

AI methods must be tied to technical solutions, not abstract descriptions of data processing.

C. Jurisdictional Variation Matters

Some jurisdictions (e.g., South Africa, possibly Australia) have been more permissive about AI inventorship. Others (U.S., Europe, UK) require human inventors but may be more flexible about patent eligibility of AI‑related technical systems.

D. Ownership and Inventorship Disputes Will Increase

As multiple contributors (scientists, programmers, data engineers) collaborate with AI, disputes like Pannu‑style inventorship corrections are likely to increase.

📌 Conclusion

Patent issues around AI‑engineered inventions such as solar‑foam cooling pavements center on inventorship, patent eligibility, and ownership. While AI can assist in discovery and optimization, **current legal frameworks in major jurisdictions require that:

A human be named as the inventor.
The invention must be more than an abstract idea (especially in U.S. patent law).
Patent applications must clearly document human contribution and technical advantages.**

The evolving case law — from DABUS challenges to evolving USPTO guidance and eligibility tests — shows that courts and patent offices are adapting but have not yet fully embraced AI as an inventor.