In re Google LLC, No. 26-157 (Fed. Cir. Oct. 6, 2026)

PER CURIAM.

O R D E R

Google LLC petitions for a writ of mandamus seeking to compel transfer from the United States District Court for the Western District of Texas (“WDTX”) to the United States District Court for the Northern District of California (“NDCA”). The University of Southern California (“USC”) opposes.¹ For the following reasons, we grant the petition.

2USC filed this suit in WDTX accusing Google, headquartered in NDCA, of infringing U.S. Patent Nos. 8,206,929 and 8,264,504 based on certain functionality of Google Earth, Google Maps, and Google Street View. Google moved to transfer the case to NDCA under 28 U.S.C. § 1404(a), arguing the parties, the named inventor, physical and electronic evidence, and the majority of witnesses relevant to the accused functionality were located in or near NDCA and that WDTX had no meaningful connection to the case. The district court found that the willing witness factor slightly favored transfer and the rest of the factors were neutral. It ultimately denied the motion.

Our review of transfer rulings is governed by the law of the regional circuit, here, the Fifth Circuit. See In re TS Tech USA Corp., 551 F.3d 1315, 1319 (Fed. Cir. 2008). Section 1404(a) authorizes a court to transfer a civil action “[f]or the convenience of parties and witnesses, in the interest of justice[.]” The Fifth Circuit has made clear that a motion to transfer “should” be granted if “the movant demonstrates that the transferee venue is clearly more convenient.” In re Volkswagen of Am., Inc., 545 F.3d 304, 315 (5th Cir. 2008) (en banc). When a district court’s denial-of-transfer ruling amounts to a clear abuse of discretion that produces a patently erroneous result, mandamus relief is appropriate. Id. at 312; TS Tech, 551 F.3d at 1319.

We conclude that the denial of transfer here was patently erroneous. In reaching that conclusion, the district court found no transfer factor favored keeping the claims against Google in WDTX. On the other hand, the court found that transfer was supported by perhaps the most important factor on this record: the convenience of the witnesses. See In re Genentech, Inc., 566 F.3d 1338, 1343 (Fed. Cir. 2009). The district court found the sources of proof, local interest, and the compulsory process factors were all neutral as to transfer. But all those factors appear to likewise point towards NDCA here.

3There are at least two problems with the conclusion that the sources of proof factor here favors neither forum. One problem with that conclusion is that the only specific sources of proof discussed by the district court in its transfer analysis are in NDCA, not WDTX. In particular, the district court discussed prior art software maintained on a CD kept in NDCA. The district court erred in giving the location of that CD no weight based on the view that the software could be loaded onto a computer and analyzed outside NDCA. See Volkswagen, 545 F.3d at 316 (“That access to some sources of proof presents a lesser inconvenience now than it might have absent recent developments does not render this factor superfluous.”).

Another problem with weighing the sources of proof factor as neutral here is that it is inconsistent with the court’s analysis of another closely related factor. The district court considered the location of potential employee witnesses as proxies for the location of potential sources of proof. But if, as the district court here reasonably explained, the willing witnesses factor had at least some weight in favor of transfer because the key technical employee witnesses are in or closer to NDCA while primarily sales and marketing employees are in WDTX, Appx7, then, by the district court’s own logic, the sources of proof factor should also at least have some weight in favor of transfer.

The district court also incorrectly found the localized interest factor neutral. The uncontroverted evidence indicated that Google designed, developed, and implemented the accused functionality in NDCA and not at all in WDTX, giving NDCA a significant interest in this litigation. See In re Samsung Elecs. Co., 2 F.4th 1371, 1380 (Fed. Cir. 2021) (noting the location where the accused functionality was “researched, designated, and developed” is a “significant factor[]” for the local interest analysis); In re Apple Inc., 979 F.3d 1332, 1345 (Fed. Cir. 2020). USC’s allegations concerning willful infringement also appear to center on events that took place in NDCA and potential witnesses4 to those events who still remain in that area. See Appx52; 2437; see also In re TikTok, Inc., 85 F.4th 352, 364 (5th Cir. 2023) (noting the place of the alleged wrong is an important consideration relevant to this factor).

USC says it filed this suit in WDTX because Google maintains a significant presence in WDTX tied to the accused products. Appx2419. But the only local connection found below was that Google performed quality assurance testing on one of the accused products in WDTX. Appx8–9. And it appears undisputed that testing had no connection to the accused functionality and general testing of the accused products occurred in NDCA. Appx2480, Appx2599. While this provides WDTX a local interest, it is not comparable to that of NDCA. See Def. Distrib. v. Bruck, 30 F.4th 414, 435–36 (5th Cir. 2022) (noting the focus of the local interest factor “most notably regards not merely the parties’ significant connections to each forum writ large, but rather the significant connections between a particular venue and the events that gave rise to a suit” (cleaned up)).

The district court weighed the compulsory process factor as neutral on the basis that “multiple witnesses are subject to the subpoena power of WDTX and NDCA.” Appx6. That analysis, however, fails to account for the fact that it appears undisputed that seven potential non-party witnesses are subject to compulsory process in NDCA, while only five are in WDTX. Of particular significance, one of the potential witnesses in NDCA is the inventor, whose testimony may be critical to the events underlying this litigation. Two other individuals in NDCA are former Google employees that USC indicates were connected with the inventor’s work. Appx79, 89, 572, 711. By contrast, USC pointed to former Google employees with knowledge of the accused products generally rather than of the accused features. Appx2426.

Even without disturbing the district court’s conclusion as to the compulsory process factor, the decision to deny5 transfer here was patently erroneous. At bottom, this case is about a California university suing a Northern California company for products designed and developed in Northern California based on patents invented by a resident of Northern California. Several factors (willing witnesses, sources of proof, and local interest) favor transfer and no factor disfavors transfer. See In re Toyota Motor Corp., 747 F.3d 1338, 1341 (Fed. Cir. 2014) (granting mandamus to compel transfer where “nothing favors the transferor forum, whereas several factors favor the transferee forum”). “In these circumstances, the district court’s no-transfer conclusion was a clear abuse of discretion.” Id.

Accordingly,

IT IS ORDERED THAT:

The petition is granted. The order denying transfer is vacated and the district court is directed to grant Google’s motion to the extent that the case is transferred to the United States District Court for the Northern District of California under 28 U.S.C. § 1404(a).

FOR THE COURT

October 6, 2026 Date