By Michael Rosen
While legislative efforts at patent eligibility reform, introduced to great fanfare two years ago, have largely stalled, public comments in response to a request by the United States Patent and Trademark Office (USPTO) have revived the issue in light of recent developments in artificial intelligence (AI).
Eligibility refers to what technological subject matter may be patentable in the first place, as longstanding precedent forbids patenting laws of nature, natural phenomena, and abstract ideas. The landmark 2014 CLS Bank v. Alice Supreme Court case tightened eligibility rules and flummoxed inventors, scholars, judges, and practitioners alike. Over the years, the USPTO has issued guidelines designed to provide greater clarity in the post-Alice eligibility sphere, and in 2019, Sens. Chris Coons (D-DE) and Thom Tillis (R-NC) introduced legislation that would have loosened the Supreme Court’s strictures. But that measure foundered on the shoals of the 2020 election, the pandemic, and other concerns.
Nevertheless, in March, Coons and Tillis, along with Sens.
Tom Cotton (R-AR) and Mazie Hirono (D-HI), sent a letter to acting USPTO director Drew Hirshfeld requesting
he publish a request for information (RFI) on the “state of patent eligibility
jurisprudence in the United States,” then evaluate and summarize the RFI’s
findings. The senators noted they were especially interested to learn how the
current jurisprudence affects “investment and innovation in critical
technologies like quantum computing, artificial intelligence, precision
medicine, diagnostic methods, and pharmaceutical treatments.”
In July, Hirshfeld duly published the RFI, soliciting input from a variety of stakeholders. And
last month, the public comment period closed, revealing a range of interesting
responses from companies, academics, and patent practitioners.
On one hand, organizations that depend on patents as their lifeblood exhibited serious concern over the current regime. The International Business Machines Corporation (IBM) lamented “significant uncertainty for businesses that innovate in the field of information technology” caused by the unsettled nature of eligibility issues. And it asserted that “current patent eligibility jurisprudence negatively impacts many aspects of IBM’s acquisition and exercise of intellectual property rights, such as patent application development, procurement, enforcement, licensing, and litigation.”
Specifically, with respect to AI inventions, IBM wrote that the Alice regime is “biased against computer-related inventions because abstraction is a foundational characteristic of computer science.” This tendency, the company contended, puts AI inventions at an inherent disadvantage because applicants will be forced to narrow their claims more than necessary. Similarly, Ericsson asserted in its own letter that with AI, a “lack of clarity and uncertainty on § 101 eligibility could allow other countries to command a greater share of the market because of a void in U.S. policy.” And the 21st Century Coalition for Patent Reform, a group comprising firms like 3M, Eli Lilly, General Electric, and Johnson & Johnson, agreed, as did the American Intellectual Property Law Association, which observed that “the current state of patent eligibility jurisprudence has had a negative impact on the predictability and reliability of patent protection, particularly in the life sciences and software industries.”
But companies finding themselves on the opposite side of
patent lawsuits disagreed. Google, for instance, asserted
that the current eligibility regime “already promote[s] investment and progress
in AI and quantum computing technologies,” as “the IP system is playing its
proper role.” As evidence, Google pointed to the explosive growth in AI patent
applications even during the post-Alice years, with American companies
receiving more than 70 percent of all AI-related patents (including Google’s
own robust AI patent portfolio). Rather than fundamentally changing eligibility
law, Google “encourage[d] the PTO to take a neutral and data-driven approach
that will allow for a robust understanding of the extensive public and private
activity in these fields.”
Similarly, the Internet Association, which comprises a wide variety of software companies like Amazon, Microsoft, and Intuit, praised the Alice regime, contending that it “has encouraged patent prosecutors to file more complete disclosures and better claims directed to the inventor’s technical advance.” The group also claimed that eligibility rules have not “affected the amount spent on R&D regarding artificial intelligence or quantum computing.”
And the Electronic Frontier Foundation, which has long
sought to reduce the impact of baseless software patent claims, contended
that after Alice, “fewer bad patents were granted and litigated, while software
innovation has proceeded in leaps and bounds.”
So who is right? As usual, the truth falls somewhere in between. Yes, since Alice, AI innovation and patent applications have continued to soar, but both likely would have increased further without the tightened eligibility limits. Yes, the Supreme Court’s 2014 opinion has eliminated plenty of bad patents, but it’s also gone too far in preventing the protection of worthy ideas, including in the AI space, that are not as readily defined as we might like. It will be interesting to see how the USPTO, under the likely new leadership of nominee Kathi Vidal, threads the needle.
The post Public comments on patent eligibility elevate issues surrounding artificial intelligence appeared first on American Enterprise Institute – AEI.