INTRODUCTION 1140 I. PATENTING UNDERDEVELOPED INVENTIONS 1145 A. Claiming the Unproven 1146 B. Prophetic Examples 1148 C. The Role of Enablement 1150 II. UNDERSTANDING EXPERIMENTAL FAILURE 1152 A. The Ubiquity of Failure in science 1152 B. The File Drawer Problem. 1153 C. The Costs of Nondisclosure 1155 III. CAN AN INVENTOR CONCEAL FAILURE? 1158 A. The Patent Applicant's Duty of Disclosure 1159 1. Basic Principles 1159 2. .Breach and Inequitable Conduct 1161 3. Submitting Misleading Technical Information 1162 B. The "Inoperative Embodiments" Doctrine 1165 1. Contours 1165 2. Theoretical Rationale 1168 3. Robustness 1171 IV. CLAIMING AROUND FAILURE 1173 A. simple Claims 1174 B. Complex Claims 1175 V. DISCLOSING FAILURE IN THE PATENT RECORD 1177 A. The Post-Filing Information Disclosure Statement 1178 B. Incentivizing Full Disclosure 1179 1. Examination Perks 1180 2. Defensive Publication 1181 C. Normative Implications 1183 1. On Patent (Examination) Quality 1183 2. On Technology Progress 1185 CONCLUSION 1187 INTRODUCTION
The U.S. Patent and Trademark Office (Patent Office) is often criticized for issuing a large number of questionable, "bad," or low-quality patents. (1) Patent quality can be defined as "the capacity of a granted patent to meet (or exceed) the statutory standards of patentability--most importantly, to [cover inventions which are] novel, nonobvious, and clearly and sufficiently described." (2) Famous examples of questionable patents include a motorized ice cream cone, (3) a method of exercising a cat with a laser pointer, (4) a snake leash, (5) and a "high-five" machine.6 Aside from being technically invalid, (7) these patents impose costs on the legal system, competitors, would-be inventors, and society, (8) The quality of an issued patent depends on the quality of the underlying examination.9 The assurance of a high-quality patent examination is largely about information (10): an examiner must have all of the relevant technical information in hand in order to accurately gauge patentability. When the issue is whether the invention works as described, the question becomes one of enablement, the statutory requirement that an invention be sufficiently disclosed to allow a person having ordinary skill in the art (PHOSITA) (11) to make and use it without undue experimentation. (12) From an information standpoint, the Patent Office is at a clear disadvantage because the inventor has little incentive to disclose failure or any experimental data that might compromise patentability. (13) And since the Patent Office lacks its own testing facilities, it must rely on information presented by the inventor. (14) This information asymmetry inevitably allows nonenabled patents to slip through the cracks and further contributes to the patent quality problem. (15)
But other factors contribute to the proliferation of nonenabled patents. First, an inventor is not required to actually make everything that is claimed (or verify that everything that is claimed actually works) before filing a patent application. As discussed below, an inventor can satisfy enablement with so-called "prophetic" examples. (16) Second, for a variety of reasons, applicants have an incentive to file patent applications early in the inventive process. (17) In fields like chemistry, biotechnology, and pharmaceuticals, this means filing during the initial stages of research and development before much experimentation has been done and the level of uncertainty is high. Together, these factors set the stage for an issued patent covering subject matter that does not work as described. (18)
Of course, inventors who file early continue to experiment during the years of patent prosecution. (19) Post-filing experimentation inevitably reveals more information about the invention than the inventor knew at the time of filing, including if any of the claimed embodiments (20) fails to work as described in the patent application. This raises an important issue that has been overlooked by both courts and commentators: When post-filing experimentation during patent prosecution reveals that some of the claimed subject matter fails to work as described, is there a duty to amend the patent document or disclose the failure to the Patent Office?
To illustrate, consider the following hypothetical. Suppose an inventor at a drug company seeks to make a class of drugs to successfully treat arthritis. Having made one compound, W, which shows promising results, the inventor hypothesizes that structurally similar compounds X, Y, and Z (which have not been made) will exhibit the same or better efficacy. So the drug company files a patent application which: (l) discloses actual details about W; (2) discloses prophetic details aboutX, Y, andZ; and (3) claims W,X, Y, and Z and a method of treating arthritis by administering a therapeutically effective amount of the respective compound to a patient. But post-filing experimentation during the application's pendency (21) reveals that while W and X work as described, Y has no effect on arthritis and Z cannot be made. Thus, [GAMMA]and Z are nonenabled.
It is unclear how the drug company should proceed. The options include:
(1) amending the written description of the invention (22) to delete details about the "inoperative" embodiments FandZ; (23)
(2) amending the claim(s) to cancel the nonenabled subject matter; or
(3) remaining silent and allowing the patent (application) to issue as it was initially filed, thereby concealing the failure.
As the discussion below will show, the relevant law is not as clear-cut as one might hope. First, a patent applicant owes the Patent Office a duty of candor, a duty of good faith, and a duty to disclose all information material to patentability. (24) A breach of this duty is considered inequitable conduct or fraud on the Patent Office and renders the patent unenforceable. (25) Second, the U.S. Court of Appeals for the Federal Circuits has reiterated that "[c]laims are not enabled when, at the effective filing date of the patent, [a PHOSITA] could not practice their full scope without undue experimentation." (27) In theory, this should provide an incentive for the drug company to address the failure during patent prosecution because the claim would be vulnerable to an invalidity attack, or the defense of inequitable conduct, if it were ever asserted in patent infringement litigation. Third, notwithstanding the previous point, the Federal Circuit and its predecessor court, the U.S. Court of Customs and Patent Appeals (C.C.P.A.) (28) have held that a claim which includes some inoperative embodiments is not necessarily invalid for nonenablement. (29) This arguably provides an escape hatch for inventors who learn about failure during patent application pendency.
There is every reason to believe that concealment of inoperability is pervasive, thereby producing a countless number of issued patents which disclose and claim failure. This problem not only contributes to the proliferation of nonenabled (and thus low-quality) patents, but has far-reaching implications for the patent system. This Article argues that an applicant who learns about experimental failure during patent prosecution has a legal obligation to, at the very least, amend the claims. (30) It then explains how to encourage inventors to disclose details about the failure in the patent record. (31) This additional disclosure would have several upsides for the inventor, (32) improve patent (examination) quality,33 enrich the public storehouse of technical knowledge, (34) and promote broader goals of the patent system. (35)
The remainder of the Article proceeds as follows. Part I explores the patenting of underdeveloped inventions. It explains why the patent system allows (if not encourages) inventors to do so despite the high likelihood that some of the claimed subject matter might not work as described. Part II examines experimental failure and the costs of nondisclosure on the scientific community and the patent system. In Part III, I argue that despite the murkiness of the relevant case law, a patent applicant cannot conceal known failure without breaching a duty of candor and good faith owed to the Patent Office. Next, Part IV explains how a patent applicant who learns about experimental failure after filing can claim around it. Finally, Part V explains how to encourage inventors to disclose information about failure in the patent record. It explores why incentivizing such post-filing disclosures has several upsides for the patent system, including improved patent (examination) quality and furthering the broader goal of promoting technological progress.
PATENTING UNDERDEVELOPED INVENTIONS
While all would agree that issuing nonenabled patents is far from ideal, their proliferation is not surprising given that "the patent law[s] place strong pressure on filing the patent application early in the development of the technology, often before... all of the boundaries [are] fully explored." (36) Indeed, inventors must often file before actually reducing the invention to practice in order to attract investors, (37) minimize risk, (38) and to safeguard patent rights in the United States and abroad. (39)
Nevertheless, the pressure to file early essentially invites inventors to seek patents on underdeveloped inventions. This regime inevitably produces nonenabled patents which disclose and claim failure.
Claiming the Unproven
An inventor can obtain a patent without actually making and testing everything that is claimed. (40) It is well settled in U.S. patent law that the mental act of conception of the idea, rather than any physical act, is the important facet of the inventive process. (41) Thus, an applicant who "constructively" reduces an invention to practice by filing a patent application which describes the invention presumably has complied with the disclosure requirements of 35 U.S.C. [section] 112(a), including...