Using a CDN May Contribute to Finding Personal Jurisdiction–R18 v. ThisAV

The plaintiff runs R18, a Japanese-focused adult site. It has US copyright registrations for 50k+ videos. The defendant runs ThisAV, which the court describes as a Japanese-language video-hosting service. When I visited ThisAV (NSFW) this morning, it self-described (via Google Translate): “ThisAV is an all Chinese adult porn video and photo album sharing community site.” The site appeared to flip between Chinese and Japanese language presentation, but that could have been a Google Translate issue.

ThisAV users allegedly republished 13 of R18’s videos. R18 sued ThisAV in the Western District of Washington for copyright infringement. ThisAV responded that the court didn’t have personal jurisdiction. ThisAV is based in Hong Kong, and one of its directors lives in Canada. The evidence connecting ThisAV to the US:

  • It uses a hosting company with servers in Utah.
  • It procured CDN services from Cloudflare for the North America and Asia markets.
  • It had English-language compliance disclosures, including pages specific to 512/DMCA and 2257, plus this odd statement: “[] is a website available from its location in the United States of America. We at do not warrant or make any discretion about its appropriate use and availability outside the aforementioned country.”
  • It sells its ad inventory to third-party services, some of whom geolocate ads to the US.
  • 4.6% of its pageviews came from the US, a total of 1.3M views.

The Ninth Circuit applied the Calder Effects Test, which imposes jurisdiction if the defendant “(1) committed an intentional act, (2) expressly aimed at the forum state, (3) causing harm that the defendant knows is likely to be suffered in the forum state.”

Intentional Act. “YMP operated, and…Lee purchased its domain name and domain privacy services.” (Huh? That’s true in every case.)

Express Aiming at the US. “the advertising structure Defendants employed demonstrated that they profited from viewers in the United States market” because it had 1.3M US pageviews. (Huh? Also true in every case.) I would have liked to see the panel discuss UMG Recordings v. Kurbanov, which specifically analyzed geolocated ads.

The panel also said that ThisAV appealed to US users, despite the fact that it was primarily in the Japanese language. The supporting evidence:

  • “by choosing to host in Utah and to purchase CDN services for North America, Defendants chose to have the site load faster for viewers in the United States and slower for viewers in other places around the world. Given how important loading speed is to achieving and maintaining an audience, Defendants’ choice is good evidence that they were motivated to appeal to viewers in the United States more than any other geographical location.” OK, some obvious problems here. First, North America ≠ US. It’s insulting to the rest of North America to equate the two. (Note: there’s about 120k Japanese in Canada and 75k Japanese in Mexico). Second, improving speed in North America doesn’t slow down the speed in the rest of the world. At most, it means that CDN turbocharges the download speed in its markets, but other markets still get the standard speed. So the panel’s factual assertion is just flat-out wrong. Third, ThisAV correctly argued that North American CDN services will improve latency globally, not just in the putatively targeted markets. The panel responds that it has to accept all plaintiff factual assertions as true.
  • “the webpages Lee and YMP posted on the site that address legal compliance are relevant almost exclusively to viewers in the United States.”

While the panel botched the CDN discussion, the panel has an on-point discourse about the disadvantages of slow page loading:

The time it takes for a site to load, sometimes referred to as a site’s “latency,” is critical to a website’s success. For one, swift loading is essential to getting users in the door. The faster a site loads in a particular location, the better its search engine optimization (SEO) will be there, that is, the higher it will appear in search engine results as compared to competitors. Appearing early in search results is critical to attracting web traffic. Studies show that 93% of online experiences begin with the user typing something into a search engine, and that the results on the first page receive nearly 92% of overall traffic for that term. Swift loading is also crucial to keeping potential site visitors engaged. Research shows that sites lose up to 10% of potential visitors for every additional second a site takes to load, and that 53% of visitors will simply navigate away from a page that takes longer than three seconds to load. Even tiny differences in load time can matter. Amazon recently found that every 100 milliseconds of latency cost it 1% in sales.

I hope the Ninth Circuit will keep these issues in mind when it hears the appeal of the challenge to AB 2273, which interposes a massive barrier to user access to websites. If a little latency costs websites money, imagine what a privacy-invasive and time-consuming age authentication access restriction will do.

Causing Harm in the US. “The operators of actively appealed to a U.S. audience, knew that a significant number of people in the United States were actually viewing the website, and were put on notice that they were hosting infringing content when Will Co. sent them a takedown notice. In light of that, it’s hard to see how Defendants could have failed to anticipate the harm that occurred in the forum.”

As a result, the panel sends the case back to the district court to consider the rest of the personal jurisdiction analysis.


Overall, this opinion leaves a lot to be desired. The panel’s vagueness about the specific jurisdictional problems attributable to ads and CDNs gives plaintiffs a lot to chew on, while the panel could have relied more on the server location and US-centric compliance disclosures to keep things cleaner. More likely, the panel should have ruled for the defense, especially in light of the fact that the service was Japanese-language and therefore had limited appeal to US audiences. I think that would have been more consistent with the Wanat precedent.

Given the factual error about CDNs, I imagine the panel will issue a corrected opinion if the defense asks for a rehearing. But this could also be a good case for a rehearing en banc to better harmonize this ruling with the Wanat and Mavrix cases.

Structurally, the panel also did not fully connect the copyright infringement claims to the alleged US ties. In other words, it’s not just that ThisAV has US contacts; the jurisdiction analysis should link those ties to the infringements, which isn’t so easy to do given that the infringements were allegedly committed by users, not ThisAV itself. The panel set up the evidence to make the connections clearer; it just didn’t do the analysis.

Case Citation: Will Co., Ltd. v. Lee, 2022 WL 3906205 (9th Cir. Aug. 31, 2022)