What Ex Post Review Has Revealed About Patents. Purpose of the Project

Transcription

1 What Ex Post Review Has Revealed About Patents Duke Law Center for Innovation Policy Roundtable on the PTAB s Post Grant Review Proceedings: A Review of the Evidence Saurabh Vishnubhakat (presenting) Associate Professor of Law, Texas A&M University School of Law (2015 ) Fellow, Duke Law Center for Innovation Policy David L. Schwartz Professor of Law, IIT Chicago Kent College of Law Alan C. Marco Chief Economist, United States Patent and Trademark Office This research represents the individual efforts of the authors and should not be imputed to the USPTO or to any other organization. This research is based entirely on data that is publicly available. Purpose of the Project Empirically estimate what IPR and CBM reveal about the patents that are challenged Compare the review that patents currently undergo in these procedures with the review that the same patents initially underwent during examination Compare the characteristics of patents that are undergoing IPR or CBM review with the characteristics of contemporary patents that have not (yet) undergone IPR or CBM review 1

2 Distinguishable Outcomes We may find several effects at work, and patent claims in IPR or CBM may turn out to be: 1. Valid 2. Invalid (new prior art) 3. Invalid (old unconsidered prior art) 4. Invalid (same prior art, new law) 5. Invalid (same prior art, same law) Working Hypotheses For decisions to institute CBM reviews: Institution differential (grant% deny%) may be higher for 101 challenges than for prior art based 102, 103 challenges 101 has been in flux more fer le ground CBM prior art limited: what was available pre AIA But KSR has raised the nonobviousness standard, so 103 is also more powerful than it used to be 2

3 Working Hypotheses For decisions to institute IPRs: Institution differential (grant% deny%) should be higher for 103 challenges than for 102 challenges 103 challenges more likely to benefit from new prior art Data We examine bibliographic characteristics of patents and patent prosecution histories IPR and CBM: public data from USPTO Examination: publicly available file wrappers of patents challenged under IPR or CBM Data collection is ongoing The following results are from a subset of the IPR and CBM data for which some prosecution data has also been collected 3

4 Initial Results Perhaps unsurprisingly, CBM reviews appear to be a forum mostly for 101 challenges Petitions made under 102, 103, and 112 are as likely as not to be instituted (p > 0.05) Petitions made under 101 are nearly a split in favor of institution (p < 0.05) 4

5 Initial Results Meanwhile, IPRs are more for 103 challenges Petitions made under 102 are as likely as not to be instituted (p > 0.05) Petitions made under 103 are more than a split in favor of institution (p < 0.05) 5

6 Initial Results Usage of IPRs appears to be technology specific Over half the decisions so far to institute or not have pertained to Computers & Communications related patents 6

7 Institution Decisions Initial Results For most technologies, petitions are instituted more often than not by about a margin (p < 0.05) For mechanical inventions, they are equally likely (p > 0.05) 7

8 Institution Decisions Initial Results These technology specific differences between institutions and non institutions are driven more by 103 than by 102 For 102 petitions, institutions are about equal to non institutions (p > 0.05) in most technologies Except Computers & Communications (p < 0.05) 8

9 Institution Decisions Initial Results For 103 petitions, institutions significantly outpace non institutions (p < 0.05) for all technologies except Mechanical 9

10 What Should IPR/CBM Correct? If IPRs and CBM reviews are for catching mistakes that USPTO examination allowed, such as false negatives in grounds for rejection Then we should expect that patents for which IPR or CBM reviews are instituted differ from patents for which reviews are not instituted Particularly with respect to grounds for rejection What Is IPR/CBM Review Correcting? Suppose Patent 123 did not initially receive any 103 rejections during prosecution, but is now challenged on 103 grounds Suppose Patent 456 did receive a 103 rejection during prosecution (and overcame Then we should expect that IPR or CBM review will be more likely to be instituted for Patent 123 than for Patent

11 What Is IPR/CBM Review Correcting? Instead, the success rate for instituting IPR and CBM petitions appears to be the same for patents that were previously rejected on a certain basis during examination and for patents not previously rejected on that basis (p > 0.05 for all grounds examined) 11

12 What Is IPR/CBM Review Correcting? So arguing new grounds in IPR/CBM petitions (grounds that did not previously come up during examination) does not make it any more likely that the petition will be instituted This may suggest that intervening changes in legal standards are playing a greater role than previously expected Next Steps Compare prior art used in examination with prior art used in IPR/CBM review Rejection during examination on the same grounds as the grounds for the current challenge is one thing But knowing whether the same prior art was involved then as now adds additional necessary insight about examination Continue updating IPR/CBM data as well as prosecution history data for relevant patents 12

13 COMMENTS & QUESTIONS 13