Sean Seymore on Utility

In patent cases, the term “hindsight bias” refers to fact-finders’ tendency to use their knowledge of the invention at issue in their analysis of whether that invention would have been obvious. This error occurs when fact-finders ignore the rule that obviousness of a claimed invention must be evaluated at the time of patent filing rather that at the time of litigation. Professor Sean Seymore’s latest article, Foresight Bias in Patent Law, deals with an error that implicates the future rather than the past. Seymore is concerned with the utility requirement, which denies patents on chemical compounds that lack a demonstrated consumer end use, such as a therapeutic use, and on methods of making such compounds. Foresight Bias builds on Seymore’s earlier article, Making Patents Useful, which criticizes the utility requirement for being too subjective and calls for its elimination. Seymore’s work may be contrasted to that of Professor Michael Risch, who sees a greater role for the utility requirement (see also here).

Seymore criticizes courts for denying patents on microscale building blocks, like chemical intermediates and gene fragments, but not on macroscale building blocks, like bricks. He argues that the two types of building blocks are similar in that they facilitate further development and are thus useful to the artisans in the field—chemists and masons, respectively. Although Seymore does not completely reject the possibility that patent law can justifiably develop technology-specific standards and acknowledges that chemistry and biotechnology are fairly characterized as “unpredictable” fields, he argues that “courts should not craft technology-specific rules based on speculation about the potential negative consequences of granting a patent.” In Seymore’s view, this is the crux of the problem: Patents on chemical intermediates have been wrongly denied based on “hypothesized negative effects of upstream patents on downstream research” that have not been empirically proven and, for some types of patents, possibly disproven. Instead of utility, which he considers too blunt of a tool against patents on chemical intermediates, Seymore would address preemption concerns on a case-by-case basis by applying the requirements of enablement and nonobviousness to constrain the scope of such claims and to ensure technical merit. Seymore’s approach would discourage overbroad claiming by mandating stringent disclosure requirements and a clear demonstration of how the claimed invention constitutes a nonobvious improvement over the prior art.

 

Although fact-intensive inquiries that would be required under enablement and obviousness standards would increase administrative costs relative to the current approach, Seymore’s proposal has much to recommend it.  Because claims to chemical intermediates and methods of making them can differ quite widely depending on the core structure of the chemical product and the number of embodiments covered, some claims are surely a greater threat to downstream research than others. The utility doctrine as currently implemented does nothing to distinguish between these different claims. Moreover, under the current regime, a patentee might obtain a composition claim entitling it to control all of a compound’s future uses upon demonstrating even a trivial utility—so long as that utility is “specific,” or not widely shared among chemical compounds generally. This example suggests that utility’s role in protecting downstream innovation from patents is questionable, and Seymore concludes that it is a “superfluous patentability requirement which does no real work.”

Nonetheless, utility may be drawn to a somewhat different concern than enablement and nonobviousness. One wonders whether the policy against the patenting of artifacts of basic research provides a basis for a requirement of patentability that cannot be adequately enforced with enablement’s “undue experimentation” standard or the nonobviousness standard. Under this approach, utility may be understood to play a “backstop” role similar to that arguably played by the patentable subject matter requirement, which can bar a patent even when the inventor taught the world how to practice the full scope of the claim and made a nonobvious improvement over the prior art. This justification does not answer the critique about technology specificity—indeed, I tend to agree with Seymore that it is odd for a requirement of patentability to apply in only one field. Perhaps, the utility requirement should cover other artifacts of basic research, not just chemical or biotechnological ones. If so, we are left with some questions. How does one identify a patent on basic research? Is complete elimination of such patents even a good idea? Like hindsight bias, foresight bias and overpessimism about effects of certain patents on future research may be very difficult to avoid. Perhaps, however, guideposts can be developed for identifying situations where patent-based incentives for the creation and commercialization of an invention will be generally outweighed by harmful effects of a patent on that invention on downstream innovation—no matter what the field.

Note: Cross-posted on the Center for Law and the Biosciences Blog