### AI Can Help Invent, But Can’t Take the Credit: Unpacking the New USPTO Guidance

For years, a question has loomed over the world of innovation: if an artificial intelligence creates something new and useful, who gets the patent? The AI itself? Its programmer? The user who prompted it? The U.S. Patent and Trademark Office (USPTO) has finally weighed in, releasing landmark guidance that draws a clear line in the sand: AI can be a crucial tool in the inventive process, but it cannot be named an inventor on a patent.

This long-awaited clarification aims to balance the explosive growth of AI-powered research and development with the long-standing legal framework of intellectual property. The core of the guidance rests on the interpretation of the Patent Act, which requires an “inventor” to be a human being.

#### The Human Element Remains Essential

The USPTO’s position is that while an AI system can contribute to the conception of an invention, a human must have made a “significant contribution” to be named as the inventor. The guidance essentially treats AI as an incredibly advanced tool, akin to a sophisticated computer program or a specialized laboratory instrument. The tool doesn’t get the credit; the person who skillfully wields it does.

This decision directly addresses cases like the attempt to list the AI system “DABUS” as the sole inventor on patent applications, a move that was rejected in the United States, the UK, and Europe. The new rules codify this stance, ensuring that inventorship remains a human endeavor.

#### What Counts as a “Significant Contribution”?

The crucial question for innovators and patent attorneys is what, precisely, constitutes a “significant contribution” in the age of AI. The USPTO provides several examples of human involvement that would likely meet the threshold:

* **Constructing the Prompt:** A person who designs a specific prompt or series of prompts to guide the AI toward a particular solution is contributing significantly to the final concept.
* **Recognizing the Output’s Significance:** An AI might generate thousands of potential solutions, but it is the human who recognizes a specific output as valuable, useful, and inventive. This act of recognition and insight is a key part of the inventive process.
* **Making a Key Decision:** The person who decides to pursue a specific AI-generated idea and refines it for a practical application is a contributor.
* **Refining the AI’s Output:** Taking a raw idea from an AI and adapting, testing, and modifying it into a workable invention is a clear inventive step.

Conversely, simply owning an AI, pushing a button, or providing a general, high-level goal without further involvement is not enough to claim inventorship. The human must be an active participant in the conception of the invention.

#### The Duty of Disclosure

A critical part of the new guidance is the requirement of transparency. Patent applicants now have a duty to disclose to the USPTO if and how AI was used to create the invention described in their application. This allows patent examiners to properly assess whether the human inventorship requirement has been met. Failing to disclose the use of AI could put the validity of any resulting patent at risk.

#### The Path Forward

The USPTO’s guidance provides much-needed clarity for companies and researchers who are increasingly integrating AI into their workflows. It allows them to confidently invest in AI-driven innovation without the legal uncertainty that previously clouded the patent process.

While AI itself cannot hold a patent, its role as a powerful partner in human creativity is now officially recognized. The future of invention isn’t about humans versus machines; it’s about the synergy between them. The credit, and the patent, however, will remain firmly in human hands.