After LKQ: The Boilerplate Changed; The Rejection Rate Did Not

by Dennis Crouch

In a recent post, I shared a chart showing that the §103 rejection rate for design patent applications has stayed flat in the two years since LKQ Corp. v. GM Global Technology Operations LLC, 102 F.4th 1280 (Fed. Cir. 2024) (en banc). Through Hague System disposals, only about 1% of design applications receive an obviousness rejection. That is roughly where the rate sat a decade ago, and a long way from what one might expect given the Federal Circuit's decision enabling more rigorous obviousness analysis. I wanted to dig into a handful of cases to see any difference from pre- and post-LKQ obviousness rejections.  Examiners have adapted their language and are also using prior art that would not have worked with the old Rosen framework, just not on a very aggressive basis.  It may be, as readers suggested, that examiners would do more if given more time to search.

For this study, I compared the text of about 150 pre-LKQ obviousness rejections (2023) against about 150 post-LKQ obviousness rejections (January 2025 through April 2026).  As you might expect, examiners are no longer citing Rosen or other pre-LKQ decisions.  80% of pre-LKQ case cited to Rosen or an equivalent case, with 0% of post-LKQ rejections. "Basically the same," the verbal marker of the Rosen primary-reference test and the Durling-era phrase "so related" both almost entirely fell off from the rejection text.

What replaced the old framework were a set of post-LKQ phrases: "visually similar" (0% → 90%) and "overall appearance" (40% → 75%) now dominate. Oddly, LKQ itself rarely cited (only 4% of rejections) and KSR is at 0%. Almost 90% of post-LKQ §103 rejections cite no obviousness precedent at all.

I worked through several rejections in detail to see whether examiners are pushing into prior art that the old framework would have blocked. Some are. Aristocrat's slot-machine GUI was rejected over a structurally different primary, and TOPPAN's curved-glass panel was rejected over a utility patent combined with four NPL web URLs.  But, the most aggressive single-reference stretch that I saw in the dataset is actually pre-LKQ: Apple's over-ear headphone rejected under Rosen-Durling over a Monster Cable in-ear earbud.

---

To continue reading, become a Patently-O member. Already a member? Simply log in to access the full post.