Tari Labs, LLC v. Lightning Labs, Inc.

Opinion

TARI LABS, LLC, Case No. 3:22-cv-07789-WHO

Plaintiff, ORDER ENTERING TEMPORARY v. RESTRAINING ORDER

LIGHTNING LABS, INC., Re: Dkt. No. 25 Defendant.

Tari Labs, LLC, brings this motion for a temporary restraining order (“TRO”) against Lightning Labs, Inc., seeking to temporarily restrain Lightning’s use of the name “TARO” for its protocol because it infringes Tari’s trademark for the “TARI” protocol. Both are blockchain- based software protocols used to create and transfer digital assets, and both companies market their protocols to similar software developer customers. Although Tari purports to make consumer-facing products itself while Lightning creates the protocol so outside developers can make consumer-facing products, those retail products seem to overlap both in use and in customer bases, and so the similar names and marks are likely to confuse customers. For those and the following reasons, on this record I find Tari is likely to succeed on the merits, has shown likelihood of irreparable harm, and shows the balance of equities and public interest favor a TRO. Tari alleges in this trademark infringement lawsuit that Lightning’s use of the name “TARO” for its blockchain-based “protocol” infringes Tari’s trademark for its “platform” of blockchain and cryptocurrency-related products, including its own blockchain-based protocol, “TARI.”1 The “TARI protocol” is an “open-source” software based on blockchain technology that “enables users to create and transfer digital assets.” Declaration of Naveen Jain (“Jain Decl.”) [Dkt. No. 28] ¶ 1. The protocol was initially launched in April 2020 by a team of individuals experienced in cryptocurrency and startups. Id. ¶¶ 9-10. Though still in its development, the TARI protocol will allow users to transfer many kinds of digital assets, “including stablecoins, non-fungible tokens (NFTs), tickets, in-app purchases on mobile devices, loyalty points, and other assets.” Id. ¶ 2. Tari targets developers for the use and development of the protocol itself, and also targets retail consumers with other products and services from its TARI platform, that use or work with the Tari protocol. Id. ¶¶ 2-4. Its products currently include a “Tari wallet” and “Tari tokens,” though it plans to expand. See id. ¶¶ 4-6. Tari and the TARI protocol have been featured by various news sources and have been marketed to consumers, and Tari has a moderate social media following. Id. ¶¶ 10-12. The TARO protocol was developed after the Bitcoin blockchain was updated in late 2021 with functionality called “Taproot,” which allows for more types of transactions to take place on that blockchain.2 Declaration of JP Singh (“Singh Decl.”) [Dkt. No. 38] ¶ 30; Declaration of Olaoluwa Osuntokon (“Osuntokon Decl.”) [Dkt. No. 36] ¶¶ 15-17. In response to this development, Lightning—which builds tools intended for developers to use to build consumer- facing applications—announced in April 2022 the “TARO protocol,” which is built using the Taproot functionality. Jain Decl. ¶¶ 27-29, 31. The TARO protocol is intended to be used to issue and transfer digital assets over the blockchain. Id. ¶¶ 31-33. Lightning says the TARO protocol 1 The trademark is for: Cryptocurrency trading and exchange services, namely, providing a digital currency or digital token for use by members of an on-line community via a global computer network; cryptocurrency trading and exchange services, namely, providing a digital currency or digital token, incorporating cryptographic protocols, used to operate and build applications and blockchains on a decentralized computer platform and as a method of payment for goods and services and as a method of transfer of digital assets. Declaration of Naveen Jain (“Jain Decl.”) [Dkt. No. 28] Ex. 8. was named as such because one founder liked the name, because another founder was Nigerian and ate taro as a child,3 and because “taproot” and “taro” are related. Declaration of Elizabeth Stark (“Stark Decl.”) [Dkt. No. 39] ¶ 39; Osuntokon Decl. ¶¶ 21, 25. Concerned that that TARO protocol would operate in the same product area and confuse customers with its name choice, Tari executives reached out to Lightning executives several times throughout 2022, alerting them to the TARO trademark and concerns about consumer confusion and asking for a name change. See Stark Decl. Ex. 16; Declaration of J. Noah Hagey (“Hagey Decl.”) [Dkt. No. 26]; Supplemental Declaration of Naveen Jain (“Jain Supp. Decl.”) [Dkt. No. 45] Ex. 1. Tari contends that the TARI and TARO protocols directly compete by selling their products and by seeking feedback from similar developers. See Jain Decl. ¶¶ 20-23. Tari asserts that both companies use the same marketing channels, media and technology platforms, and industry press publications. Id. ¶¶ 24-26. Tari also says that Lightning’s TARO protocol is receiving increased press attention, in part due to recent “Community Calls” Lightning has held to promote the product. Id. ¶¶ 28-30. Lightning ignored or denied the accusations throughout 2022, see Stark Decl. Ex. 16; Hagey Decl.; Jain Supp. Decl. Ex. 1, and in September, it publicly released its open-source “TARO protocol.”4 Singh Decl. ¶ 33. In December 2022, Tari sued for trademark infringement. [Dkt. No. 1]. Tari now asserts that on February 5, 2023, Lightning’s Chief Technology Officer announced that a new version of the TARO protocol would be released within a month. See Jain Decl. ¶ 32. Tari says that the following day, another Lightning executive stated that this newest version of the TARO protocol would have “just about everything needed to fully get off the ground.” Id. ¶ 3; see also id. Ex. 27. Concerned this impending “launch” would immediately harm its brand, reputation, and business, Tari filed the present motion for a temporary restraining order (“TRO”). 3 See also infra n.11. II. PROCEDURAL BACKGROUND Tari filed its complaint on December 8, 2022. [Dkt. No. 1]. Lightning answered on February 7, 2023. [Dkt. No. 24]. On February 21, 2023, Tari filed its TRO, (“Mot.”) [Dkt. No. 25], along with nearly 500 pages of exhibits, [Dkt. Nos. 26, 27, 28]. I held a short hearing to set a briefing schedule and issued an order maintaining the status quo. [Dkt. Nos. 32, 34]. Lightning filed its opposition on February 27, 2023, (“Oppo.”) [Dkt. No. 35], along with nearly 700 pages of exhibits, [Dkt. Nos. 36, 37, 38]. Tari replied on March 2, 2023, (“Repl.”) [Dkt. No. 435], with over 150 pages of exhibits, [Dkt. Nos. 43, 44, 45]. I order Lightning to file a sur-reply specifically addressing the new evidence raised by Tari in its reply concerning the “Taro Wallet,” which Lightning did. (“Surrepl.”) [Dkt. No. 48]. I held a hearing on March 8, 2023, at which counsel for both parties appeared. LEGAL STANDARD Federal Rule of Civil Procedure 65 governs preliminary injunctions and temporary restraining orders. Fed. R. Civ. Proc. 65. The standard for both forms of relief is the same. See Stulhbarg Int’l Sales Co. v. John D. Brush & Co., 240 F.3d 832, 839 n.7 (9th Cir. 2001). A preliminary injunction may be issued if a plaintiff establishes: (1) likelihood of success on the merits; (2) likelihood of irreparable harm in the absence of preliminary relief; (3) that the balance of equities tips in his favor; and (4) that an injunction is in the public interest. See Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). “Injunctive relief [is] an extraordinary remedy that may only be awarded upon a clear showing that the plaintiff is entitled to such relief.” Id. at 22. The Ninth Circuit has held that “‘serious questions going to the merits’ and a hardship balance that tips sharply toward the plaintiff can support issuance of an injunction, assuming the other two elements of the Winter test are also met.” All. for the Wild Rockies v. Cottrell, 632 F.3d 1127, 1132 (9th Cir. 2011). “A stronger showing of one element may offset a weaker showing of another. For example, a stronger showing of irreparable harm to plaintiff might offset a lesser showing of likelihood of success on the merits.” Id. at 1131.

5 I consider and cite to the timely filed reply at Dkt. No. 43, in part because Tari did not explain I address the four factors governing TROs below. Tari makes a strong showing of success on the merits and meets the statutory rule for a rebuttable presumption of irreparable harm, which Lightning fails to rebut. The balance of equities and public interest factors are less strong but at this stage, favor Tari. Accordingly, and for the following reasons, Tari’s motion is granted. The parties fervently debate Tari’s likelihood of success on the merits of its trademark infringement claims but agree on the legal test that applies. “To prevail on its Lanham Act trademark claim, a plaintiff must prove: (1) that it has a protectible ownership interest in the mark;6 and (2) that the defendant’s use of the mark is likely to cause consumer confusion.” Rearden LLC v. Rearden Com., Inc., 683 F.3d 1190, 1202 (9th Cir. 2012) (quotation marks omitted) (citing Network Automation, Inc. v. Advanced Sys. Concepts, Inc., 638 F.3d 1137, 1144 (9th Cir. 2011)). “More precisely . . . at the preliminary injunction [or TRO] stage, [the plaintiff] must establish this it is likely to be able to show such a likelihood of confusion.” Brookfield Commc’ns, Inc. v. W. Coast Ent. Corp., 174 F.3d 1036, 1053 n.15 (citation omitted). Consumer confusion may be shown as either forward or reverse confusion. Ironhawk Techs., Inc. v. Dropbox, Inc., 2 F.4th 1150, 1159 (9th Cir. 2021). “Forward confusion occurs when consumers believe that goods bearing the junior mark came from, or were sponsored by, the senior mark holder. By contrast, reverse confusion occurs when consumers dealing with the senior mark holder believe that they are doing business with the junior one.” Id. at 1159-60 (citations and internal quotation marks omitted).7 The likelihood of confusion element is government by the eight “Sleekcraft factors”: 6 Though Lightning seems to contest Tari’s protected ownership interest, Oppo. 7 n.2, I find that Tari is likely to succeed on the merits of this element because Tari submitted evidence of its registered trademark. See Pom Wonderful LLC v. Hubbard, 775 F.3d 1118, 1124 (9th Cir. 2014) (“Registration of a mark is prima facie evidence of the validity of the mark, the registrant’s ownership of the mark, and the registrant’s exclusive right to use the mark in connection with the goods specified in the registration.” (citing 15 U.S.C. § 1115(a))). As Lightning notes, additional discovery will be necessary to contest this evidence. “(1) strength of the protected mark; (2) proximity and relatedness of the goods; (3) type of goods and the degree of consumer care; (4) similarity of the protected mark and the allegedly infringing mark; (5) marketing channel convergence; (6) evidence of actual consumer confusion; (7) defendant’s intent in selecting the allegedly infringing mark; and (8) likelihood of product expansion.” Pom Wonderful LLC v. Hubbard, 775 F.3d 1118, 1125 (9th Cir. 2014) (citing AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348-49 (9th Cir. 1979), abrogated on other grounds by Mattel Inc. v. Walking Mountain Prods., 353 F.3d 792, 810 n.19 (9th Cir. 2003)). “[T]his multi- factor approach must be applied in flexible fashion” and the facts are “not a rote checklist”; “[a] determination may rest on only those factors that are most pertinent to the particular case . . . and other variables besides the enumerated factors should also be taken into account based on the particular circumstances.” Rearden, 683 F.3d at 1209 (citations omitted). “[A] plaintiff need not satisfy every factor, provided that strong showings are made with respect to some of them.” Pom Wonderful, 775 F.3d at 1125 (quoting Surfvivor Media, Inc. v. Survivor Prods., 406 F.3d 625, 631 (9th Cir. 2005)). And as the Ninth Circuit cautioned in 1999 in a case concerning Internet domain names—a new technology at the time—courts “must be acutely aware of excessing rigidity when applying the [Sleekcraft factors] in the Internet context; emerging technologies require a flexible approach.” Brookfield, 174 F.3d at 1054 (emphasis added). A. Strength of the Mark “The stronger a mark—meaning the more likely it is to be remembered and associated in the public mind with the mark’s owner—the greater the protection it is accorded by the trademark laws.” Network Automation, 638 F.3d at 1149 (quoting Brookfield, 174 F.3d at 1058). “A mark’s strength is ‘evaluated in terms of its conceptual strength and commercial strength.’” JL Beverage Co., LLC v. Jim Beam Brands Co., 828 F.3d 1098, 1106 (9th Cir. 2016) (citation omitted). “Conceptual strength involves classification of a mark ‘along a spectrum of generally increasing inherent distinctiveness as generic, descriptive, suggestive, arbitrary, or fanciful.’” Network Automation, 638 F.3d at 1149 (quoting Brookfield, 174 F.3d at 1058). “Arbitrary or fanciful marks deserve wide protection because the trademark holder can properly expect to run into very make consumers associate the trademark with the product.” Dreamwerks Prod. Grp., Inc. v. SKG Studio, 142 F.3d 1127, 1130 n.7 (9th Cir. 1998). “Commercial strength is based on ‘actual marketplace recognition,’ and thus ‘advertising expenditures can transform a suggestive mark into a strong mark.’” Network Automation, 638 F.3d at 1149 (quoting Brookfield, 174 F.3d at 1058). In cases of reverse confusion, I assess “whether consumers doing business with the senior user might mistakenly believe that they are dealing with the junior user,” which requires “evaluat[ing] the conceptual strength of [the plaintiff’s] mark and compar[ing] it to the commercial strength of [the defendant’s] mark.” Ironhawk, 2 F.4th at 1162 (quoting JL Beverage Co., 828 F.3d at 1107). “[I]n a reverse confusion claim, a plaintiff with a commercially weak mark is more likely to prevail than a plaintiff with a stronger mark, and this is particularly true when the plaintiff’s weaker mark is pitted against a defendant with a far stronger mark.” Id. at 1163 (quoting Cohn v. Petsmart, Inc., 281 F.3d 837, 841 (9th Cir. 2002) (per curiam)). Indeed, the central question is whether the junior mark is so commercially strong as to overtake or “swamp the reputation” of the senior mark. Id. at 1162-63 (first citing Walter v. Mattel, Inc., 210 F.3d 1108, 1111 n.2 (9th Cir. 2000); and then citing Dreamwerks, 142 F.3d at 1130 n.5). The parties agree that Tari is a conceptually strong mark because it is an “arbitrary mark[].” Oppo. 9 n.3; Mot. 11:10-21; see also Dreamwerks, 142 F.3d at 1130 n.7. Tari has an “Arabic or Italian” etymology “relating to a Mediterranean gold coin used in the Middle Ages.” Mot. 11:10-12 (citing Jain Decl. ¶ 14). Though apparently chosen because coins are tangentially related to payment technology, typical consumers are unlikely to understand this association. I agree that the mark is arbitrary or fanciful. Turning to the commercial strength of the marks, Tari seems to argue both that its mark is strong and that its mark is weak. See Mot. 11:22-28 (asserting Tari has substantially invested in its brand), Mot. 12:1-8 (arguing for application of the reverse confusion doctrine, which applies where “plaintiff’s weaker mark is pitted against a defendant with a far stronger mark” (citation omitted)). Lightning contends that the TARI mark is commercially weak, Oppo. 8:9-9:3, and I tend to agree on this record—Tari has pointed to a few news articles, a moderate social media conferences. In contrast, Lightning has a commercially strong mark. Lightning notes that “many” public developers already know about and use its TARO protocol, Oppo 32:19, and Tari submits evidence of at least one developer using the TARO brand name on its newly developed wallet product, Repl. 1:10-2:3. Additionally, Tari submits evidence showing Lightning is a well-known brand with powerful financial investors, Jain Decl. ¶¶ 15-16, TARO has received significant press attention including from Forbes, id. ¶ 30, Ex. 9, and TARO has a strong social media presence, id. ¶ 31. Though unlike Ironhawk, 2 F.4th at 1163, where there was evidence of the defendant’s significant marketing expenditures related to the allegedly infringing product, here Lightning has the commercially stronger mark because of the evidence showing “actual marketplace recognition,” Network Automation, 638 F.3d at 1149 (citation omitted). Accordingly, like the plaintiff in Ironhawk, here Tari may be able to show that Lightning’s use of the TARO mark may “swamp the reputation” of the TARI mark, and Tari’s reputation and business could be affected. Based on this somewhat limited record, then, I find that Tari is likely to show that this element favors finding infringement. B. Proximity and Relatedness of the Goods The parties fervently disagree on the proximity of the goods in the market. Lightning even contests the fundamental basis of Tari’s expert report because the methodology was grounded in the assumption that the goods are proximate in the market. “Goods and services are related when they are complementary, sold to the same class of purchasers, or similar in use and function.” Ironhawk, 2 F.4th at 1163 (citing Sleekcraft, 599 F.2d at 350). A plaintiff suing for trademark infringement “need not establish that the parties are direct competitors to satisfy the proximity or relatedness factor.” Rearden, 683 F.3d at 1212. Rather, “[r]elated goods (or services) are those ‘which would be reasonably thought by the buying public to come from the same source if sold under the same mark.’” Ironhawk, 2 F.4th at 1163 (quoting Sleekcraft, 599 F.2d at 348 n.10). Within the Ninth Circuit, courts “adopt[] a rather flexible approach to the whole notion of competition.” Rearden, 683 F.3d at 1212 (collecting cases). ‘complementary’ or ‘related’ that the public is likely to be confused as to the source of the services.” Am. Int’l Grp., Inc. v. Am. Int’l Bank, 926 F.2d 829, 832 (9th Cir. 1991) (citations omitted). Lightning describes its TARO protocol as a “software protocol” that allows developers “to build new digital assets on the Bitcoin blockchain and transact with them over the Bitcoin blockchain or the Lightning Network.” Oppo. 3:22-25 (citing Stark Decl. ¶¶ 36, 41). Tari describes its TARI protocol as “a blockchain protocol . . . that allows users to create and transfer digital assets on the blockchain. . . . [and] allows developers and other market participants to build the protocol into their technology and create their own tools for transferring digital assets.” Mot. 2:19-22 (citing Jain Decl. ¶¶ 2-7). Though Lightning seems to disagree, both companies assert that they make a blockchain protocol that allows developers or other users to build and transfer digital assets. Accordingly, the protocols are “similar in use or function,” Ironhawk, 2 F.4th at 1163 (citation omitted), and the businesses are “related,” Am. Int’l Grp., 926 F.2d at 832. Lightning repeatedly contends that the TARO protocol is intended only for developers, which would mean that it is not sold to the same class of purchasers as the TARI protocol. See, e.g., Oppo. 1:11-16, 3:17-18, 4:8-11, 17:9-11. But Tari says its protocol is also directed to developers, see, e.g., Jain Decl. ¶¶ 2-3, so evidently there is at least some overlap in the customer bases. Lightning also says that it can only be used on the Bitcoin blockchain and that Bitcoin developers rarely work on other blockchains, see Singh Decl. ¶¶ 22-24, but Tari contests whether developers are so dedicated to a single blockchain, see Jain Supp. Decl. ¶ 19, points out that Lightning’s executives are encouraging overlap with technology on different blockchains, id. ¶ 22; id. Exs. 5, 6; and adds that Tari has announced plans to make its protocol compatible with the Bitcoin blockchain, which is where the TARO protocol operates, id. ¶ 20.8 8 Lightning strongly emphasizes the importance of the two protocols operating on different blockchains. As noted, it seems there is more overlap than Lightning believes. But I am also not so sure that operating on different blockchains is as important as Lightning makes it seem, so long as the products (protocols) are similar and operate in the same blockchain space. Here is a hypothetical example comparison, using a more established blockchain-based product: If Bitcoin owned a trademark protecting its famous name and cryptocurrency coin, and a new company Accordingly, there is sufficient evidence in the record to show that Tari is likely to succeed in showing that the protocols are related and proximate in the market. It is true that Lightning may be able to show, after further discovery, that the way developers and consumers encounter the product is wholly distinct—but it has not done this yet, and I am not so convinced it will be able to sufficiently in a way that defeats Tari’s assertions. This factor favors a TRO. C. Type of Goods and the Degree of Consumer Care This factor considers “the sophistication of the customers,” including whether a “‘reasonably prudent consumer’ would take the time to distinguish between the two product lines.” Ironhawk, 2 F.4th at 1167 (citation omitted). “‘When the buyer has expertise in the field,’ or ‘the goods are expensive, the buyer can be expected to exercise greater care in his purchases.’” Id. (citation omitted). It is not clear whether this factor favors Tari or Lightning given the parties’ disagreement over the relevant products. If only the protocols were at issue, this factor leans toward Lightning. The protocols are directed at software developers, and it makes sense that they are sophisticated consumers who exercise great care when choosing their software. Of course, neither party presents any evidence of whether software developers pay for the protocol—indeed, both say their protocols are “open-source,” Mot. 18:22-24; Oppo. 2:22-27; Repl. 12:7, which seems to mean accessible for free—and if the consumers are not paying for the product, of course they would be less discerning in their choices than those paying any monetary fee. But overall, at this point Tari does not sufficiently rebut that software developers take care in their software decisions. That said, less-sophisticated consumers can use and purchase the products that are built with the protocols. For example, Tari points to the wallets: Even if TARI and TARO wallets currently work on different blockchains, it is not clear from the record that an ordinary consumer would exercise sufficient care to confirm the underlying blockchain when she chooses a wallet. of its name for a nearly identical product, despite the fact that they operate on separate blockchains. That is essentially Tari’s argument here, concerning protocol instead of coins. See also Telegram Messenger Inc. v. Lantah, LLC, No. 18-CV-02811-CRB, 2018 WL 3753748, at *1- 2, 6 (N.D. Cal. Aug. 8, 2018), aff’d, 782 F. App’x 528 (9th Cir. 2019) (enjoining the use of an 1 And while Lightning argues the “TARO wallet” is meant for developers, not consumers, its 2 argument that the TARO name will never be seen by consumers is implausible, at least on this 3 record. 4 Accordingly, at this stage, I find this factor neutral. 5 D. Similarity of the Marks 6 “The following axioms define and delimit the similarity analysis: (1) similarity is best 7 evaluated by appearance, sound, and meaning; (2) marks should be considered in their entirety and 8 as they appear in the marketplace; and (3) similarities weigh more heavily than differences.” Pom 9 Wonderful, 775 F.3d at 1127 (citation and footnote omitted); see also Network Automation, 638 10 F.3d at 1151 (emphasizing the importance of addressing how consumers would encounter the 11 mark in the marketplace). “[T]he more similar the marks in terms of appearance, sound, and 12 meaning, the greater the likelihood of confusion.” Brookfield, 174 F.3d at 1055 (holding 13 “MovieBuff’ and “MovieBuff.com” were “essentially identical”). “[A]s the similarities between 14 two marks increase, so too does the likelihood of confusion.” Pom Wonderful, 775 F.3d at 1127 3 15 (citation omitted). “Non-visual similarities are particularly important where . . . there is no 16 evidence that consumers encounter the marks side-by-side in the marketplace.” Jd. at 1129 n.9. 3 17 Tari’s expert declaration submitted a screenshot of the “landing page” for each company, 18 showing Tari’s mark and Lightning’s allegedly infringing mark: 19 \} TARI Taro 21 22 Palmatier Decl. Jf] 22, 41; see also Network Automation, 638 F.3d at 1151 (noting the importance 23 of considering how the marks will be encountered “in the marketplace”). 24 The appearance and sound of the products’ marks are extremely similar. Both use the 25 same three first letters, have two syllables, and are pronounced almost identically, with both 26 emphasizing the first syllable more strongly than the second (TAR-ee; TAR-oh). See Pom 27 Wonderful, 775 F.3d at 1129 (emphasizing importance of the marks’ pronunciation); Brookfield, 28 174 F.3d at 1055 (same). Of course, all the letters in TARI are capitalized while only the “T” in

Taro is (though it seems all letters in TARO are capitalized when referencing the protocol, as referred to by the parties in their briefs), the Tari mark has a small logo next to it, and the last letter of each differs. The words’ meanings also differ. But “similarities weigh more heavily than differences.” Pom Wonderful, 775 F.3d at 1127. These marks are very similar. I am also persuaded by Tari’s point that developer-consumers of the Taro protocol are likely to use the Taro name on their consumer-facing products, in order to show consumers that the products they are purchasing are connected to the protocol. See Network Automation, 638 F.3d at 1151 (noting the importance of considering how the marks will be encountered “in the marketplace”). This is significant because it emphasizes the importance of the word itself, and therefore the similarities between the words “Tari” and “Taro,” more than the color, size, or font of the text. See iCall, Inc. v. Tribair, Inc., No. C-12-2406 EMC, 2012 WL 5878389, at *7 (N.D. Cal. Nov. 21, 2012) (“[O]ne feature of a mark may be more significant and it is proper to give greater force and effect to that dominant feature.” (quoting McCarthy on Trademarks & Unfair Competition § 23:42)); see also Pom Wonderful, 775 F.3d at 1129-30 (holding that even where “the presentation of the marks differ[ed] in terms of prominence, font, size, and capitalization,” the district court erred in finding similarity weighed against the plaintiff, because the marks had identical pronunciation, meaning, and letters). For example, the “Taro Wallet” shows the word “Taro” in white, apparently using a slightly different font from the one used by Lightning on Taro’s “landing page,” but the name “Taro Wallet” is obviously very similar to “Tari Wallet,” which is precisely Tari’s concern. See Repl. 1:16-22.9 Accordingly, I find that Tari has shown a likelihood of success concerning the similarity of marks.10

9 This also addresses a point made by Lightning’s expert, that TARO does not have a “logo” and that the image provided in Tari’s declaration is merely “a heading” on a developer page. Butler Decl. ¶ 25. Even if true, this does not address the fact that consumers are likely to encounter the “TARO” name—regardless of what the text looks like—if TARO’s non-Lightning developers continue to use the TARO to connect their products to the TARO protocol.

10 Lightning’s cited cases are not to the contrary. See Oppo. 12:22-13:2. In iCall, 2012 WL E. Marketing Channel Convergence “In assessing marketing channel convergence, courts consider whether the parties’ customer bases overlap and how the parties advertise and market their products.” Pom Wonderful, 775 F.3d at 1130 (citing Nutri/System, Inc. v. Con–Stan Indus., Inc., 809 F.2d 601, 606 (9th Cir. 1987)). “Marketing channels can converge even when different submarkets are involved so long as ‘the general class of . . . purchasers exposed to the products overlap.’” Id. (citing Sleekcraft, 599 F.2d at 353). Lightning does not contest that both parties use the same marketing channels, including “Twitter, Substack, and Github.” Oppo. 17:16-19 (agreeing Lightning also uses those marketing channels and asserting that “all companies in the software development and cryptocurrency spaces use these common online platforms”). But Lightning contends this factor does not favor Tari because their consumer bases do not

dissimilarity between the “dominant component[s]” of the words, “i” and “Wi.” He thoughtfully reasoned that while pronounced similarly (“eye” and “why”), the aural difference from adding the “w” was “not insignificant”; the word “i” had no meaning while “wi” referred to “wifi”; and most importantly, the defendants had provided evidence showing the marks were only encountered along with the “very distinct” logos. Id. Here, while the different meanings of Taro and Tari favor the defendants, the words have near-identical pronunciations and the aural difference is insignificant. And while “Tari” sometimes appears next to its logo, sometimes it does not, and “Taro” never appears with a logo; Judge Chen’s point was that the logos themselves offered distinction, which is not possible when sometimes there is no logo. See id. In addition to being out of circuit and out of district, FieldTurf USA Inc. v. TenCate Thiolon Middle E., 945 F. Supp. 2d 1379, 1390 (N.D. Ga. 2013), combined the “similarity” and “customer care” factors when it reasoned that “Evolution” and “Revolution” differed by only one letter and were pronounced differently, but customers were sophisticated and unlikely to be deceived by this difference. The order later noted that the customer bases were also distinct because the products represented entirely different stages of manufacturing. Id. I find this reasoning inapposite, and if anything, supportive of Tari because it emphasized the similarity between the names. Lastly, Lightning cites Haven Capital Management, Inc. v. Havens Advisors, L.L.C., 965 F. Supp. 528 (S.D.N.Y. 1997), aff’d, 159 F.3d 1346 (2d Cir. 1998). In addition to being out of circuit, Haven explained that “[o]bviously” the words “haven” and “havens” were similar, which favors Tari’s position. Id. at 531. Additionally, there the defendant-investment advisers were legally precluded from advertising generally, whereas the plaintiff could advertise to the public, so the defendant’s mark was only presented to a marketplace of “sophisticated investors” who were “required to put up significant money” that would be “tied up for a minimum of eighteen months.” Id. Here, Lightning repeatedly argues that its developer-consumers are highly sophisticated but overlap. Oppo. 17:6-15. I am not convinced. Both companies target their products at software developers. Jain Decl. ¶¶ 2-4, 8, 11; Singh Decl. ¶¶ 27, 31, 34. Lightning says TARO is only used by Bitcoin developers, whom Lightning asserts do not use any other blockchain technology, Singh Decl. ¶¶ 23-24, but Tari strongly contests that developers restrict themselves to a single blockchain, Jain Supp. Decl. ¶ 49, asserts that the TARI protocol will be developed for use on multiple blockchains, id. ¶¶ 20-21, and says Lightning itself is advocating for more technology to be used on multiple blockchains, id. ¶ 22. Additionally, and with respect to Lightning’s repeated assertions that its products and name will never be consumer-facing, see, e.g., Oppo 4:8-11 (“[O]rdinary consumers . . . are highly unlikely to come into contact with the TARO name.”), Tari provided evidence that Taro’s name is already being used by a developer using Taro’s protocol, for a “Taro Wallet,” Jain Supp. Decl. Ex. 13. At this stage it seems reasonable that the consumers for the Tari Wallet and the Taro Wallet are likely to overlap despite Lightning’s arguments to the contrary. At the very least, Tari makes a strong showing that “the general class of . . . purchasers exposed to the products overlap.” Pom Wonderful, 775 F.3d at 1130 (citation omitted). Accordingly, given the uncontested overlap in marketing channels, and the clear showing of likely overlap between users of the parties’ products, this favor strongly favors Tari. F. Evidence of Actual Confusion “[E]vidence of actual confusion, at least on the part of an appreciable portion of the actual consuming public, constitutes strong support for a ‘likelihood of confusion’ finding.” Reardon, 683 F.3d at 1210 (citing Playboy Enters., Inc. v. Netscape Comm’cs Corp., 354 F.3d 1020, 1026 (9th Cir. 2004)). “[C]onfusion on the part of potential consumers may be relevant” and “non- consumer confusion can serve as a proxy for consumer confusion.” Id. at 1215 (citations omitted). “Because of the difficulty in garnering such evidence . . . the failure to prove instances of actual confusion is not dispositive.” Ironhawk, 2. F.4th at 1165 (quoting Sleekcraft, 599 F.2d at 353). “Therefore, ‘this factor is weighed heavily only when there is evidence of past confusion or, perhaps, when the particular circumstances indicate such evidence should have been available.’” Id. (quoting Sleekcraft, 599 F.2d at 353). confusion on the social media application Telegram and in a news article; (2) its expert report; and (3) Lightning’s counsel’s mix-up of the marks. First, Tari’s users mistyping “Tari” and instead typing “Taro” on Telegram shows that the name is similar, can easily be mistyped, and might be mistaken inadvertently due to typographical error. See Jain Supp. Decl. Ex. 12. This favors Tari. But I am less convinced that these screenshots show confusion because it seems as though the consumers were strictly discussing Tari and all other members of the conversation understood the typo to refer to Tari, and because nothing shows any consumers knew about or were directed to the TARO protocol. Tari’s exhibit showing a mix-up in a blog post is somewhat more convincing because it demonstrates repeated use of “TARI” when the author was describing the “TARO” protocol. See Palmatier Supp. Rep. Ex. 1 at 9. While it is again clear from context that the post discusses TARO, the post appears to be an informational blog for developers or consumers, so even small typos could create confusion. See Reardon, 683 F.3d at 1215 (considering confusion on behalf of potential consumers and non- consumers as relevant to actual confusion). This somewhat favors Tari. Second, Tari’s expert report purports to show very high levels of consumer confusion. See Palmatier Rep. Lightning contends the entire report should be disregarded, and some of its expert’s criticisms seem valid, particularly regarding the lack of controls and the use of leading questions. See Butler Rep. However, Lightning did not submit an opposing report showing, for example, a survey that demonstrated no confusion among consumers (likely due to the time- limited nature of the motion and briefing schedule, though perhaps also because it is not clear whether Lightning must disprove actual confusion). And, one of its most fervent criticisms is that Tari’s expert’s use of a Squirt survey is flawed because the products do not appear in close proximity in the marketplace. See id. ¶¶ 17-27. As discussed, I am not persuaded by this argument. Accordingly, at this stage I find the survey may show the possibility of actual confusion, but I cannot say at this time that it affirmatively shows actual confusion. Third, Tari points out the mistake made by Lightning’s former counsel in responding to Tari’s demand letter, where counsel wrote “Taro Labs” instead of “Tari Labs.” See Hagey Decl. Reardon, 683 F.3d at 1215, and it is particularly poignant that Lightning’s counsel made this mistake in response to a demand letter asking its client to stop using the name out of fear of confusing customers. But I do not find this evidence is as helpful for Tari as Tari seems to believe, because it does not show the impact on actual consumers. Finally, Lightning argues that crypto investors and developers are sophisticated and unlikely to confuse the products, but in reply Tari points to information showing confusion among crypto investors concerning a recent blockchain product launch by the company Coinbase. See Repl. 12:19-25; 13:15-19 (explaining that Coinbase recently launched a blockchain network called Base, and in response crypto investors rapidly brought significant amounts of BASE, an unaffiliated and unrelated crypto token); see also Jain Supp. Ex. 10. This supports Tari’s assertion that developers and other users of cryptocurrency can be confused by similar product names, but again does not affirmatively demonstrate there was any actual confusion here. Accordingly, while this factor may tip slightly in Tari’s favor, it is “not dispositive” of the inquiry into likelihood of confusion. Ironhawk, 2. F.4th at 1165. G. Defendant’s Intent in Selecting the Mark When assessing the defendant’s intent to infringe, “[t]his factor favors the plaintiff where the alleged infringer adopted his mark with knowledge, actual or constructive, that it was another’s trademark.” Id. at 1167 (internal quotation marks omitted) (citing JL Beverage, 828 F.3d at 1111- 12). For reverse confusion cases, the court “ask[s] whether there is some evidence that the junior user, when it knew of the senior user, was at fault for not adequately respecting the rights of the senior user.” Id. at 1167-68 (quoting 4 McCarthy on Trademarks and Unfair Competition § 23:10 (5th ed. 2020)). “Intent can be shown through . . . ‘evidence that, for example, the [junior user] knew of the mark, should have known of the mark, . . . or otherwise culpably disregarded the risk of reverse confusion.’” Id. at 1168 (citation omitted). Tari repeatedly reached out to Lightning, alerted it to Tari’s trademark and its expectations and fears of confusion, and even offered to help develop a new mark. See Repl. 4:25-5:14 (collecting sources). These documents are sufficient to show, at this stage, that Lightning knew of believe that there would be actual confusion, in part because its TARO protocol was not consumer facing like some of Tari’s products. See Stark Decl. Ex. 16 (showing message exchange between Tari and Lightning executives with the latter stating, “[I] don’t think it’s similar, it’s a root vegetable” as well as “and [TARO is] not consumer-facing anyway”). But it is not clear that a belief of noninfringement—without additional confirmation or any other steps, in the face of repeated requests to change including via a demand letter, and in the context of obviously similar names—constitutes a “good-faith” belief sufficient to overcome the evidence that Lightning knew its name choice might infringe but went ahead anyway. See Ironhawk, 2 F.4th at 1167-68; cf. JL Beverage Co., LLC v. Jim Bean Brands Co., 815 F. App’x 110, 112-13 (9th Cir. 2020) (unpublished) (holding it was not clearly erroneous for district court to find intent factor neutral where defendant knew of the mark but the marks were not similar). Lightning explains the history and reasoning behind its choice of the name “Taro,” including its connection to the new Bitcoin functionality called Taproot, its personal meaning for one of its founders, and its relationship to some of the countries that Lightning hopes to work with in the future.11 Stark Decl. ¶ 39; Oppo. 3:22-4:4. But regardless of how creative or applicable the name is for the brand, none of these reasons overcome Tari’s showing—at this stage—that Lightning intentionally disregarded the risk of reverse confusion by using the name Taro, especially after repeated warnings. See Ironhawk, 2 F.4th at 1168. Accordingly, this factor favors Tari. H. Likelihood of Product Expansion Even if I were to assume these protocols were “non-competing goods,”12 a “‘strong possibility’ that either party may expand [its] business to compete with the other will weigh in 11 At the hearing, counsel for Lightning strongly emphasized the role that Osuntokon’s childhood experience eating taro played in the name choice. From Lightning’s declarations, though, it is apparent that Stark came up with and proposed the name because she liked it, and Osuntokon later offered the connection to his childhood. See Stark Decl. ¶ 39; see also id. (confirming “the fact that taro happened to be a root vegetable” and so connected to Taproot was not initially considered in the naming decision).

12 As noted, both companies make protocols for transferring digital assets on blockchain favor of finding that the present use is infringing.” Ironhawk, 2 F.4th at 1168 (quoting Sleekcraft, 599 F.2d at 354). In its reply, Tari points to a recently released “Taro Wallet” that it says strongly risks confusing consumers who use or might use its own “Tari Wallet,” and shows that Lightning is likely to expand its business to compete with Tari. Repl. 1:12-2:3. Lightning spends much of its opposition (and also much of its Answer and its initial Letter concerning the TRO hearing, see Dkt. Nos. 24, 30) insisting that its product is significantly different from Tari’s, that it is merely a protocol for use by developers, that consumers will likely never see the name “Taro,” that the protocols operate on different blockchains, and that there is no risk of confusion for those reasons. And Lightning further asserts in its sur-reply that the “Taro Wallet” is not a consumer-facing product but currently directly solely at other developers. But, Tari’s entire point seems to be that Lightning markets its protocol to developers to use to make new digital assets, and so the protocol itself and those assets bear the Taro name and therefore risk confusing consumers. Lightning wants to use the TARO name for its protocol, launch it into the world, and then turn a blind eye to the inevitable use its protocol’s name by developers making consumer-facing products that compete directly with Tari’s products. Though I understand Lightning’s assertion that it makes the protocol and does not make those other products, in this new and rapidly developing area of technology, Tari has the stronger argument at this preliminary stage.13 See Brookfield, 174 F.3d at 1054 (noting “emerging technologies require a flexible approach” (emphasis omitted)). Accordingly, Tari is likely to succeed in showing that Lightning may expand its business to compete (further) with Tari’s, which favors implementing a TRO. A plaintiff seeking a temporary restraining order or preliminary injunction for violation of 15 U.S.C. § 1125(a),(c),(d) is “entitled to a rebuttable presumption of irreparable harm . . . upon a

13 Lightning also argues that, in some respects, it literally cannot expand into Tari’s business. See finding of likelihood of success on the merits.” 15 U.S.C. § 1116(a); see also AK Futures LLC v. Boyd St. Distro, LLC, 35 F.4th 682, 694 (9th Cir. 2022) (same). Because Tari showed a likelihood of success on the merits, it is entitled to this presumption. To rebut this presumption, Lightning does not dispute that Tari’s brand will be harmed but rather asserts Tari’s delay in filing for a TRO was unreasonable. See Oppo. 20:25-23:12. It is true that Tari has known about the TARO protocol since early 2022, though it presents evidence showing it first tried to resolve the issue amicably with Lightning’s executives and subsequently filed a demand letter. Tari’s motivation for filing now was that Lightning announced it will soon release a fuller update of TARO, which will allow for more developers to use the TARO protocol and build potentially infringing products. There seems to be foundation for that concern, given the recent development of the TARO wallet and Lightning’s own concession that “many external developers have already begun building on TARO under that name.” Oppo. 23:18-19. Lightning concedes it had planned to “release” the next “version” of TARO within two months, Stark Decl. ¶ 55, but cautions that this release for open-source technology is different from a standard release for a typical consumer product, because all developments are already public, id. ¶¶ 48-57. Rather, a “release” or “launch” for open-source software is instead a “symbolic milestone[]” based on “a series of designated milestones.” Id. ¶¶ 49, 53. Accepting that Tari was indeed motivated by Lightning’s statement that it would soon release its next version of the TARO protocol, it seems the fear was less about the software being publicly released and more about the meaning of the announcement that the protocol had reached a particular milestone. In support of that concern, Tari repeatedly argues that its brand will be irreparably harmed if Lightning takes further action on its protocol, including additional announcements or “releases.” Tari presents a reasonable argument for why it waited to file until now. Based on the available evidence and presented arguments, I find that Lightning has not rebutted the presumption of irreparable harm. This factor favors a TRO. Because Lightning failed to rebut the presumption of irreparable harm, it must show that asserted harm to its reputation and brand, and that impact on potential and future customers. Lightning’s arguments about the balance of equities mirror its arguments above: that there is no evidence of confusion or harm and that Tari’s motion is untimely. See Oppo. 23:24-24:8. In contrast, Lightning asserts that it will suffer “reputational damage” from a TRO, particularly because its product is “already widely known” and because many developers have already begun using the TARO protocol. Oppo. 23:17-19. Lightning explains that restraining its public development and production has forced it to create a separate, private repository of code for its TARO protocol so that its internal team could continue developing the protocol while non- Lightning developers simultaneously worked on the open-source code that was already public. See Stark Decl. ¶¶ 101-07. It says that its reputation and brand are being harmed because it cannot actively respond to developers that contact Lightning to ask about TARO protocol updates or to submit bug reports, code, or feedback. Id. ¶ 107. First, I have doubts as to the irreparability of this asserted harm. Lightning says that the TARO protocol “has not yet ‘launched’ in any significant sense” and that it is an “early prototype” that “can be tested but is not ready for regular use.” Stark Decl. ¶ 53. Pausing the merging of its public and internal development (not pausing internal development or public development by non- Lightning users) of this early-stage product is unlikely to significantly affect the brand of a company that engages in many other ventures. Lightning also concedes that it can make up for a pause in public development by merging its internal developments with those made on its open- source code by non-Lightning users, though it says it will be “difficult and time-consuming to merge.” Stark Decl. ¶ 101. And, Tari shows that changing the name of the TARO Protocol would be relatively straightforward, see Declaration of Michael Berry (“Berry Decl.”) [Dkt. No. 44], which Lightning does not contest, though Lightning asserts that users and developers already know and work under its existing TARO name, see Stark Decl. ¶ 108. Given these considerations, it is not at all clear that Lightning will suffer irreparable harm from the maintenance of the status quo, at least until briefing for a preliminary injunction can be completed. Second, even assuming any harm is irreparable, the balance of equities can still favor the Sys. Corp. v. Se. Exp. Co., 64 F.3d 1330, 1338 (9th Cir. 1995), superseded by statute on other grounds (“Where the only hardship that the defendant will suffer is lost profits from an activity which has been shown likely to be infringing, such an argument in defense merits little equitable consideration [on an appeal from a preliminary injunction].” (internal quotation marks and citation omitted)); see also 2Die4Kourt v. Hillair Cap. Mgmt., LLC, 692 F. App’x 366, 369 (9th Cir. 2017) (unpublished) (holding the district court did not abuse its discretion by finding the balance of equities favored the plaintiffs, where the defendant’s harm of shutting down its business, terminating its employees, and defaulting on its obligations “result[ed] from [the] defendant’s allegedly infringing conduct”). Here, Lightning’s purported harms stem solely from its allegedly infringing use of TARI’s trademark—that is to say, it could continue all activities and communications if it changed the name of the TARO protocol. Accordingly, I find that the equities favor Tari, at least on this record. Lightning says the public will be harmed by a TRO because of the impact on developers working on the TARO protocol. See Oppo. 24:10-25:5. Lightning does not articulate what that impact would be but reiterates that there are many members of the public who actively contribute to the development of TARO and work on “unaffiliated” projects using TARO, and that a TRO would force those developers to work on the public open-source version while Lightning simultaneously develops its internal version. Id. Tari argues that preventing consumer confusion is in the public interest and notes (without evidence) that “the entire industry is under pressure to weed out bad actors and build community trust.” Oppo. 24:10-11. At this juncture, it is not clear that the public will be particularly affected by maintenance of the status quo. Tari cites a case from the Central District that reasons, “In trademark cases, the public interest is the right of the public not to be deceived or confused as to the source of the parties’ goods.” FAZE Apparel, LLC v. Faze Clan, Inc., No. 2:18-CV-02052-RGK-JEM, 2018 WL 3830027, at *8 (C.D. Cal. May 22, 2018) (finding the public interest weighed in favor of an injunction where the defendant’s products were likely to cause consumer confusion). I find this the TARO protocol is likely to confuse consumers. For that reason, and to avoid such confusion, the public interest favors entering a TRO. VI. FRCP 65(C) BOND REQUIREMENT “Rule 65(c) invests the district court with discretion as to the amount of security required, if any.” Johnson v. Couturier, 572 F.3d 1067, 1086 (9th Cir. 2009) (internal quotation marks and citation omitted). “[T]he likelihood of success on the merits . . . tips in favor of a minimal bond or no bond at all.” People of State of Cal. ex rel. Van De Kamp v. Tahoe Reg’l Plan. Agency, 766 F.2d 1319, 1326 (9th Cir.), amended, 775 F.2d 998 (9th Cir. 1985) (citation omitted). Because Tari established likelihood of success on the merits, and because of the speculative nature of Lightning’s financial injury from the TRO, I will not require a bond. The motion for a TRO is GRANTED. Lightning is hereby RESTRAINED from making external updates to its TARO protocol, from merging its internal updates with its public-facing open-source code, and from announcing or otherwise “launching” the next stage or “milestone” of the TARO protocol. Lightning is NOT wholesale restrained from responding to communications from non-Lightning developers and users, though it may not use those communications to further develop its internal or public code repositories. For example, Lightning may respond to inquiries about “when certain features are going to be implemented” but it may not review or substantively respond to submitted bug reports, code, or feedback. See Stark Decl. ¶ 107. This Order shall restrain Lightning until the hearing on any motion for dissolution of the TRO or for a preliminary injunction. The parties shall meet and confer on an appropriate schedule for these motions, considering what discovery would be necessary, if any, before the motions are filed. The parties should include a normal briefing schedule, providing the Court with at least two weeks after the last brief is filed prior to the hearing. In the event of disagreement over the schedule, the parties may file a Joint Statement containing the proposed schedules and no more than 2 pages each explaining the superiority of that party’s schedule. The Joint Statement shall be 1 remain in place until a hearing is held on the filed motion(s) and a decision is made. 3 Dated: March 13, 2023

5 liam H. Orrick 6 United States District Judge 4 8 9 10 11 12

© 15 16

= 17

Z 18 19 20 21 22 23 24 25 26 27 28