top of page
  • Writer's pictureCorinne Atton

The NSCAI Calls for a Review of IP Laws to Incentivize AI Innovation

In this final of three short summaries of the recent United States National Security Commission on Artificial Intelligence (“NSCAI”) Report, we identify its key messages to Congress and the White House regarding intellectual property (“IP”) law and policy in the U.S.

1. IP Policy Must be Made a National Priority

First, the NSCAI urges the U.S. government to make IP a national priority, and to reform and establish new IP policies aimed at incentivizing, expanding and protecting innovation specifically in artificial intelligence ("AI") and other “critical emerging technologies.”

2. IP Laws Should Be Reviewed to Ensure U.S. AI Innovation is Incentivized

The NSCAI identifies certainty, and a strong and robust patent system as “critical” to incentivize American innovation, and to incentivize investment in emerging technologies. Patent eligibility, in particular, is identified as “a key factor” in decisions to invest in technologies and to bring a new product to market.

The NSCAI voices concern at the current perceived restriction on the ability of innovators to obtain patents, particularly on computer-implemented and biotech related inventions, and warns that innovators may rely on trade secret protection, which prevents valuable technical information entering the public domain.

Three specific issues are flagged by the NSCAI: (1) inventorship by AI; (2) the state of the law as regards the patent eligibility and validity of AI inventions; and (3) U.S. leadership in the standard-essentials patent process.

As to inventorship, U.S. law requires a determination as to who legally contributed to the conception of an invention. Conception requires an inventor to have a specific solution to a problem. In line with the position taken in various other jurisdictions, U.S. law requires an inventor to be a natural person. See, for example, the recent United States Patent and Trademark Office (“USPTO”) DABUS decision (“a machine does not qualify as an inventor under the patent laws”). The NSCAI urges the U.S. government to assess the need for policy changes for issues raised by AI-generated inventions and creations, as technologies evolve.

As to the state of the law, following the Supreme Court decision in Alice Corp. Pty. Ltd. v. CLS Bank Int., 573 U.S. 208 (2014), a two-step test is followed when determining U.S. patent eligibility:

(1) Are the patent claims at issue directed to a patent-ineligible concept (a law of nature, natural phenomena or an abstract idea)?

(2) What else is there in the claims? Considering the elements of each claim both individually and as an ordered combination, are there additional elements that transform the nature of the claim into a patent-eligible application, such that the patent in practice amounts to significantly more than a patent on the ineligible concept itself? This has been called the “search for an ‘inventive concept’.”

The NSCAI expresses concern that the approach taken by the U.S. Court of Appeals for the Federal Circuit (the Federal Circuit) as to the application of this test, appears to diverge from that taken by the USPTO during patent prosecution, and urges the U.S. government to consider legislation to reform the patent eligibility doctrine.

As the NSCAI notes, as regards patent eligibility and the validity of AI inventions, the USPTO published a report in October 2020, outlining Public Views on AI and IP Policy. Key findings in this report include contributor reviews that:

· AI is dynamic, and the definition of AI may change over time.

· AI can be understood as computer functionality that mimics cognitive functions associated with the human mind (e.g., the ability to learn).

· AI inventions include: (1) inventions that embody an advance in the field of AI; (2) inventions that apply AI (to a field other than AI); and (3) inventions that may be produced by AI itself.

· Humans (natural persons) remain integral to the operation of AI—and AI can neither invent nor author without human intervention.

· A natural person may contribute to the conception of an AI invention in many ways and each case needs to be assessed on a case-by-case basis.

· It may be difficult for AI inventors to satisfy the U.S. patent law requirement of “enablement” (showing others how to make and use the invention without undue experimentation).

· It may be more difficult to establish the “person having ordinary skill in the art” hypothetical legal standard for AI inventions (a person who is deemed to have normal skills and knowledge in the particular technical field—a benchmark standard from which certain aspects of patent validity are assessed).

· As AI inventions proliferate, an unprecedented volume of “prior art” may be created, which may make processing and assessing AI patent applications difficult.

As to the standard-essentials patent process, the NSCAI urges intervention by the U.S. government to ensure that U.S. companies “are able to fully participate in the processes by which ‘standard essential’ patents are claimed and asserted.” The NSCAI flags concerns as to “the continuing legitimacy of the standard-setting process,” which is a privately developed method for coordinating the development and deployment of new technologies—in particular, urging a review as to whether steps are needed “to protect the integrity of international standards-setting” to prevent undue influence by foreign governments, including an over-declaration of patents as “standard essential,” and a review of any “barriers to U.S. participation in foreign standard-setting bodies.”

3. Data Protection Must Be Considered

Data is essential to AI, and the collection and compiling, in particular of “big data” (very large data sets that may be analyzed computationally to reveal patterns, trends, and associations) is very valuable. The NSCAI voices concern as to whether current U.S. laws—in particular patent and copyright laws—adequately protect data, particularly data for AI algorithms and machine learning. The NSCAI urges the U.S. government to assess whether there is a need for additional IP-type protections for AI, to ensure that investment, public disclosure and sharing are incentivized.

4. International Participation and Coordination is Essential

Finally, the NSCAI urges the U.S. government to coordinate on IP policies internationally, with the World Intellectual Property Organization (WIPO) and with other countries. Specific foci are identified to be the development of global disincentives for IP theft, and the minimization of any inconsistencies in patent regimes that hinder the ability of companies to protect their inventions in multinational markets.

(Image acknowledgement: USPTO)

bottom of page