PTAB decision note

Kayak v. IBM (CBM2016-00075): Denying Review of Art Already Considered in Prosecution

In Kayak v. IBM, CBM2016-00075, Paper 16 (December 15, 2016), the Board denied covered business method review. Its obviousness grounds relied on references examined at length during twelve years of prosecution, so the Board exercised discretion under 35 U.S.C. § 325(d) and § 324(a). It also found the § 101 challenge not shown more likely than not.

On this page
  1. Background
  2. What the Board decided
  3. Why it matters
  4. Later developments
  5. Key takeaways

Kayak Software Corp. v. International Business Machines Corp., CBM2016-00075, is an early, detailed example of the Board refusing to revisit prior art that the Office had already examined. In Paper 16, entered December 15, 2016, a three-judge panel declined to institute covered business method review of IBM’s U.S. Patent No. 7,072,849 on any ground. It is listed here with the Board’s informative decisions.

Background

The petitioners (Kayak, OpenTable, Priceline.com and The Priceline Group) challenged claims 1–25 of the ’849 patent, which concerns presenting advertising in an interactive computer network, including storing advertising objects locally at the user’s reception system to reduce network traffic. IBM had asserted the patent in district court.

The petition raised a § 101 ground and several obviousness grounds built on four references: an article the parties called “Reference 7,” Simon, Alber and Wilson. The prosecution history was unusually long. The application was filed in 1993 as a divisional of a 1989 application, and the patent issued in 2006, after multiple Office actions and three Board decisions on appeal or rehearing. Reference 7, Simon and Alber were all cited on the face of the patent. The petitioners conceded those three had been considered, but argued that the Office never considered them in the exact combination now proposed.

What the Board decided

Section 325(d)

Section 325(d) allows the Director to reject a petition because “the same or substantially the same prior art or arguments previously were presented to the Office.” The Board applied it to every obviousness ground:

  • The record could not have been clearer. The prosecution history devoted dozens of pages to each of Reference 7, Simon and Alber, and every two-reference pairing of them appeared in at least one paper. The earlier Board panel had shown it was willing to raise new references on its own.
  • A new combination was not enough. Instituting review just because the exact three-reference combination had never been stated would, the Board said, “exalt form over substance.”
  • Adding Wilson did not change the result. Wilson was new, but the petition used it only for a dependent-claim feature the examiner had already treated as unpatentable without any prior art, and Simon was cited for the same feature elsewhere.

The Board stressed that similarity of art does not require denial. Thinner prosecution, clear examiner error, or materially changed circumstances such as a new claim construction or new priority evidence could weigh toward institution, but the petitioners had identified none.

Section 324(a)

The Board separately relied on its discretion under § 324(a) and 37 C.F.R. § 42.208. The more than twelve years of prosecution, at least six prior-art Office actions and three Board decisions weighed heavily against instituting the prior-art grounds.

Section 101

On eligibility, the Board was not persuaded that the claims were directed to an abstract idea. It found the petitioners’ brick-and-mortar analogies a stretch and agreed with IBM that the patent addressed a problem arising in computer technology (bandwidth) with a computerized solution (local storage). It did not reach step two of the Alice framework, and it did not decide whether the patent qualified for CBM review.

Why it matters

Kayak shows the Board treating § 325(d) as a real filter: a petitioner cannot avoid it by rearranging or incrementally supplementing art the examiner already weighed. The decision also points to the kinds of facts that can overcome the concern. Those ideas were later developed in Becton, Dickinson & Co. v. B. Braun Melsungen AG (see the Becton Dickinson note) and in the Board’s later two-part framework for § 325(d). For the institution thresholds generally, see the different threshold standards.

Later developments

The CBM program under AIA § 18 expired for new petitions on September 16, 2020, but § 325(d) continues to apply to post-grant review, inter partes review and reexamination requests. Discretionary-denial practice has changed repeatedly, so readers should check current USPTO guidance. Designations can also change; check the USPTO’s list of precedential and informative decisions.

Key takeaways

  • In Kayak v. IBM, CBM2016-00075, Paper 16, the Board denied review on all grounds.
  • Art extensively considered during prosecution was “the same or substantially the same” under § 325(d), even in a new combination.
  • Adding a reference only for a minor dependent-claim feature did not avoid § 325(d).
  • Material errors or changed circumstances can weigh in favor of institution.
  • The § 101 ground failed at step one because the claims addressed a computer-specific problem.

Primary Sources

  1. 37 C.F.R. Part 42
  2. USPTO Patent Trial and Appeal Board (decisions)
  3. 35 U.S.C. § 324 (Institution of post-grant review)
  4. 35 U.S.C. § 325 (Relation to other proceedings or actions)
  5. America Invents Act (including § 18)

Educational information; not legal advice. Statutes, rules and USPTO practice change. Confirm against the primary sources and consult a registered patent practitioner before acting. Disclaimer.