JUSTM2J LLC v. AYDEN BREWER, JON LITZ, JASON ST. GEORGE, and JOHN DOE 1, et al.
No. 2:25-cv-0380-DAD-SCR
UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF CALIFORNIA
August 19, 2025
Before the court is a motion by Defendants Ayden Brewer, Jon Litz, and Jason St. George to stay discovery pending the District Judge‘s decision on their motions to dismiss (ECF Nos. 15-16) the First Amended Complaint (“FAC“), or, alternatively, to limit discovery to the identification of any Doe defendants. ECF Nos. 39, 44. For the reasons provided below, the Court largely denies the motion, but stays jurisdictional discovery.
I. INTRODUCTION
A. Operative Complaint
The FAC explains that the Opentensor Foundation (“Opentensor“) is the owner and developer of the Bittensor network (“Bittensor“), an open-source platform AI researchers can use to develop and improve each other‘s digital commodities. ECF No. 14 at ¶ 21. Users are paid for meaningful computations or machine-learning in TAO, digital tokens worth hundreds of dollars
California resident Brewer and New York resident St. George were Bittensor developers until early 2024, using the screennames “Rusty” and “Philanthrope,” respectively. Id. at ¶¶ 12-13. They also owned and operated Vertex Storage Solutions, LLC (“Vertex“), headquartered in Sacramento, California. Id. at ¶¶ 12-13, 43. They incorporated Vertex in December 2023 to serve as owner of the “FileTAO” Bittensor subnet. Id. at ¶¶ 12-13, 42. Litz, a Missouri resident whose online handle is “0xJones,” unsuccessfully applied to be a Bittensor developer in January 2024, after which he operated FileTAO with Brewer and St. George. Id. at ¶ 14.
Before leaving Opentensor, St. George was the developer who designed Bittensor‘s software code. Id. at ¶ 34. He was also one of only six people who had Opentensor‘s API key.1 Id. at ¶ 27. Before leaving Opentensor, Brewer previewed several elements of what would become the Bittensor Attack to other Opentensor employees. Id. at ¶ 39. This included transferring cryptocurrency without being detected, using “privacy coin” like Monero to conceal proceeds, and registering accounts on exchanges that do not require using the owner‘s legal name. Id. at ¶ 39. As of the Bittensor Attack, Litz was a member of a Telegram channel related to Railgun, a service enabling users to obscure cryptocurrency transaction information. Id. at ¶ 40.
On May 20, 2024, Defendants registered the domain name opentensor.io, specifically to mimic Opentensor‘s domain. Id. at ¶ 46. On May 22, 2024, Opentensor released a Bittensor update, “version 6.12.2,” onto Github. Id. at ¶ 47. Moments later, Defendants illicitly used Opentensor‘s PyPI key to upload their own file named “version 6.12.2” to PyPI (“Attack Software“), blocking the upload of the legitimate version 6.12.2 update. Id. at ¶ 48. The Attack Software was identical to the legitimate update on Github except that once the users made a
Also on May 22, 2024, FileTAO released an update requiring users to install “version 6.12.2” of Bittensor. Id. at ¶ 85. Requiring Bittensor users to install Bittensor updates immediately upon their release is itself unusual. Id. at ¶ 86. In any case, while those who downloaded the Bittensor update from Github were unaffected, those who used PyPI downloaded the Attack Software instead. Id. at ¶¶ 49-50. Between May 22 and July 2, when Opentensor discovered and removed the Attack Software from their PyPI account, Defendants used 32 different wallet keys to steal 61,793.90 TAO. Id. at ¶¶ 51-54.
Defendants attempted to launder the TAO via the tools and techniques Brewer had previewed to Opentensor employees. They exchanged TAO for other crypto assets before exchanging those assets for Monero. Id. at ¶ 57. They transferred assets to exchanges and private wallet addresses. Id. at ¶ 57. These wallets include Railgun and a private wallet that was used in at least one prior cryptocurrency theft and possibly a money laundering operation (“Suspected Laundering Service“). Id. at ¶ 57. Efforts to hide the stolen cryptocurrency has succeeded insofar as its current location is not completely known. Id. at ¶¶ 58-62.
Some evidence within the blockchain connects Defendants to the Bittensor Attack. For example, one of Litz‘s Ethereum wallets (“Subnet Wallets“), 0xD5, transferred assets to the Suspected Laundering Service between June 8 and July 16, 2024. Id. at ¶ 76. This transaction also used an intermediate address associate with the Bittensor Attack, 0x5E. Id. at ¶ 77.
In May 2024, before the attack, Defendants attempted to negotiate an expedited $3.5 million sale of FileTAO. Id. at ¶ 84. Defendants stopped communicating with the prospective purchaser on June 8, began transferring FileTAO‘s assets out of Bittensor on June 10, and deregistered FileTAO on Bittensor on June 11. Id. at ¶ 88. St. George and Litz deleted their online presence on Discord, Twitter, and other social media sites. Id. at ¶¶ 89-90. Vertex dissolved in September. Id. at ¶ 91.
Thirteen victims, representing 99% of the lost TAO, assigned all legal claims to Plaintiff after the Bittensor Attack. Id. at ¶ 93. Plaintiff is a Delaware entity with a principal place of
B. Procedural History
Plaintiff filed this action on January 27, 2025, and the FAC on February 28. ECF Nos. 1, 14. On April 8, Brewer moved to dismiss the action for failure to state a claim. ECF No. 15. Litz and St. George also moved to dismiss the action for lack of personal jurisdiction and failure to state a claim. ECF No. 16-18. Both motions have been fully briefed and were submitted without oral argument. ECF Nos. 24-29, 33-35, 38.
On May 27, the parties submitted a joint status report pursuant to
Defendant filed this motion to stay discovery on June 11, which the parties subsequently briefed pursuant to a joint statement. ECF Nos. 39, 44. Although Plaintiff has not served discovery requests on Defendants, Plaintiff has served subpoenas on Coinbase Global Inc. (“Coinbase“) and Rusty‘s RV Rentals LLC (“Rusty‘s“). ECF Nos. 39-3, 39-4, 44 at 4. The subpoena served on Coinbase seeks records for eight cryptocurrency addresses that Plaintiff contends were “used by wallets” indirectly “connected” to Defendants. ECF No. 39-3 at 7, 10; ECF No. 44 at 4. Coinbase paused production of responsive documents pending the outcome of this motion. ECF No. 44 at 4. The subpoena served on Rusty‘s seeks all financial records, government agency filings, negotiable instruments, rental agreements, communications, cryptocurrency-related documents, and records of investments by St. George since May 1, 2024. ECF No. 39-4 at 7, 10-11. Rusty‘s objects to the subpoena and has refused to produce any information. ECF No. 44 at 3.
////
A district court exercises “wide discretion in controlling discovery.” Little v. City of Seattle, 863 F.2d 681, 685 (9th Cir. 1988). It has similarly broad discretion to stay proceedings as an incident to its power to control its docket. See Clinton v. Jones, 520 U.S. 681, 706 (1997); see also Lockyer v. Mirant Corp., 398 F.3d 1098, 1109 (9th Cir. 2005) (court‘s power to stay proceedings is incidental to power to control cases with economy for itself, counsel, and litigants). The ordinary course of litigation is for discovery to proceed in the face of a pending dispositive motion. Courts disfavor blanket stays of discovery because “delaying or prolonging discovery can create unnecessary litigation expenses and case management problems.” Salazar v. Honest Tea, Inc., 2015 WL 6537813, at *1 (E.D. Cal. Oct. 28, 2015) (citation omitted). On the other hand, a stay of discovery pending the resolution a potentially dispositive motion may further the goal of efficiency for the courts and the litigants. See, e.g., Little, 863 F.2d at 685.
Courts in the Ninth Circuit typically rely heavily on two factors to determine if delaying discovery is appropriate: (1) whether the pending motion is potentially dispositive of the case, or at least would render unnecessary the discovery at issue; and (2) whether the pending motion can be decided absent additional discovery. See Salazar, 2015 WL 6537813, at *1. The first prong is not satisfied if disposition of the motion would likely involve leave to amend. See, e.g., Mlejnecky v. Olympus Imaging Am., Inc., 2011 WL 489743, at *9 (E.D. Cal. Feb. 7, 2011) (finding a pending motion to dismiss not dispositive of the case where the Magistrate Judge anticipated that, even if the motion were granted, the District Judge would grant leave to amend). “In applying the two-factor test, the court deciding the motion to stay must take a ‘preliminary peek’ at the merits of the pending dispositive motion to assess whether a stay is warranted.” Yamasaki v. Zicam LLC, 2021 WL 3675214, at *1 (N.D. Cal. Aug. 19, 2021) (citation omitted).
Courts both within and beyond the Ninth Circuit also regularly apply additional factors in deciding whether to stay discovery. These “Skellerup factors” include “[t]he type of motion and whether it is a challenge as a matter of law or the sufficiency of the allegations; the nature and complexity of the action; whether counterclaims and/or cross-claims have been interposed; whether some or all of the defendants join in the request for a stay; the posture or stage of the
III. ANALYSIS
Given that the parties have held their
A. PLAINTIFF ANTICIPATES TAILORING ANY REQUEST FOR JURISDICTIONAL DISCOVERY TO THE DISTRICT JUDGE‘S RULING
Plaintiff requested jurisdictional discovery in opposition to the motion to dismiss by St. George and Litz. ECF No. 26 at 18-19. However, Plaintiff noted that any “jurisdictional discovery request would necessarily be tied to the scope and basis of any determination by this Court that the jurisdictional showing set forth [in the opposition] and in the Complaint falls short[.]” Id. at 1 n.1. Plaintiff accordingly requested that it be permitted to “submit a more detailed jurisdictional discovery request” after the Court rules on the motions to dismiss. Id. at 19 & 1 n.1. By Plaintiff‘s own account, then, jurisdictional discovery should await Judge Drozd‘s
////
The undersigned recognizes the possibility that certain merits discovery will be intertwined with jurisdictional discovery.2 Such dual-purpose discovery may appropriately proceed. However, any purely jurisdictional discovery will await Judge Drozd‘s rulings and any jurisdictional discovery request Plaintiff might then make. Disputes about whether any discovery is dual purpose or purely jurisdictional may, after a meet and confer, be brought promptly before the undersigned through an informal discovery conference (“IDC“).3
B. MERITS OF THE MOTIONS TO DISMISS AND THE IMPLICATIONS FOR DISCOVERY
In ruling on the motion to stay, the undersigned must take a “preliminary peak” at the pending motions to dismiss in order to evaluate their scope and the likelihood that they will resolve the entire case. See Mlejnecky, 2011 WL 489743, at *8. “The undersigned recognizes the awkward nature of this review procedure here.” Id. The “responsibility” for “resolv[ing] defendant‘s motion to dismiss plaintiff‘s First Amended Complaint belongs only to” Judge Drozd, and Judge Drozd “may take a different view of the merits.” Id. Given that this is only a preliminary peek, the analysis below will not be as detailed or complete as it would be if the motions to dismiss were being fully adjudicated. With that caveat, the analysis proceeds.
The motions to dismiss collectively raise three types of arguments. Both motions argue the FAC generally fails to sufficiently plead specific facts as to Defendants’ wrongdoing, particularly under the heightened pleading standard for fraud. ECF No. 15 at 17-22, 28, 30-33; ECF No. 16 at 26-31. Brewer also raises several arguments as to why specific claims fail as a matter of law or fact. ECF No. 15 at 22-28, 32-33. Litz and St. George argue that the FAC does
That two motions to dismiss are pending in this action, and the motion for a stay hinges on both, reflects the complexity of this case. See ECF No. 44 at 2. This action concerns three Defendants, a Plaintiff representing thirteen alleged victims, and seven causes of action across federal and state law. Aside from this complexity itself disfavoring a stay (see Skellerup, 163 F.R.D. at 601), Defendants do not articulate whether certain discovery should be precluded by the prospect of success on particular aspects of their motions as opposed to others. At most, they concede that the Court “could order limited jurisdictional discovery” given St. George‘s and Litz‘s personal jurisdiction arguments “in the event the motion to stay is granted.” ECF No. 44 at 9 (quoting Spearman v. I Play, Inc., 2018 WL 1382349, at *2 (E.D. Cal. March 19, 2018)).4
Upon a preliminary peek at the motions to dismiss, the Court will only bar discovery if it determines that the pleadings fail such that leave to amend would be futile.
1. Defendants’ Attack on the FAC‘s Factual Allegations
Brewer argues in his motion to dismiss that the FAC fails to articulate specific and non-conclusory allegations about the role each Defendant played in an alleged conspiracy to steal TAO.5 ECF No. 15 at 18-19. He argues that the FAC mostly refers to the “Defendants” or “attackers” as a group, yet fails to demonstrate an agreement between Defendants beyond their professional relationship through Vertex. Id. at 19. Brewer reinforces this argument by asserting that because all claims sound in fraud, a complaint must plead the circumstances constituting such fraud “with particularity.” Id. at 18;
If the FAC does not already plead such facts, it pleads sufficient facts to show that Plaintiff could likely amend it to address any deficiencies. Although it often refers to “the
Brewer counters that many of the allegations supporting the accusation that Defendants orchestrated the Bittensor Attack are equally consistent with innocent activity. ECF No. 15 at 20. He cites Twombly to assert that showing “parallel conduct” is insufficient to demonstrate any agreement or conspiracy between Defendants. Id. (citing Bell Atlantic v. Twombly, 550 U.S. 544, 553-54 (2007)). While parallel conduct standing alone would be insufficient to plausibly allege wrongful conduct, Plaintiff appears to allege more than merely parallel conduct. See supra, Section I.A. “[W]hen the entire sequence of events in the complaint is considered in context, what might otherwise appear to have been coincidental parallel conduct on its own becomes ‘suggestive of illegal conduct’ and is thus sufficient to survive a motion to dismiss.” Soo Park v. Thompson, 851 F.3d 910, 928–29 (9th Cir. 2017) (quoting Twombly, 550 U.S. at 564 n.8). Moreover, pleadings need not refute other, potentially innocent explanations for alleged wrongful acts to be deemed plausible at the motion to dismiss stage—otherwise the requirement that reasonable inferences be drawn in a plaintiff‘s favor would be a dead letter. See Boquist v. Courtney, 32 F.4th 764 (9th Cir. 2024) (reversing and remanding in part because of district court‘s failure to draw all reasonable inferences in the plaintiff‘s favor).
Based on a preliminary peek, it is not clear that the FAC‘s allegations are generally deficient. But even if they are generally deficient, leave to amend would likely be granted to allow Plaintiff to attempt to plead additional facts.
After general arguments regarding the sufficiency of the pleadings, Brewer raises questions of both law and fact to challenge individual causes of action. Questions of law include whether Bittensor Attack victims could assign their rights to sue under the CFAA and Wiretapping Act (ECF No. 15 at 22-23, 27), to what extent an assignee can sue for punitive damages under the California Penal Code (id. at 34), whether cryptocurrency like TAO is a cognizable loss under the CFAA (id. at 23-25), whether the victims’ electronic wallets constituted “protected computers” under the CFAA (id. at 26-27), whether the Attack Software constituted a “device” under the Wiretapping Act (id. at 29), and whether the remedy of constructive trust can provide basis for a separate cause of action (id. at 33). By Brewer‘s own admission, some of these issues have not been definitively settled in Defendants’ favor. See id. at 29. Even if Brewer were to succeed on all of these purely legal arguments, the case would proceed because it is premised on both diversity jurisdiction and federal question jurisdiction. ECF No. 14 at ¶¶ 16-17.
The remaining arguments concern the sufficiency of pleadings as to specific claims. As to the claim for fraud, for example, Brewer argues that the FAC does not allege he made any knowingly false representation with the intent to defraud, or that Plaintiff sustained damages for reasonably relying on such representations. ECF No. 15 at 31 (citing Herrejon v. Ocwen Loan Servicing, LLC, 980 F.Supp.2d 1186, 1202 (E.D. Cal. 2013)). The FAC alleges that the named Defendants, as FileTAO operators, induced Bittensor users to download the Attack Software by disguising it as a Bittensor update purportedly needed to run FileTAO. ECF No. 14 at ¶¶ 12-14, 42, 48-50, 85-86. This allowed Defendants to allegedly use the Attack Software to steal wallet keys and subsequently drain these wallets of all TAO. Id. at ¶¶ 51-54. If this does not currently satisfy the pleading requirements for fraud, it at least suggests Plaintiff could amend the pleadings to adequately address any deficiencies as to Defendants’ mental states.
Brewer then argues that the conversion and unjust enrichment claims should be dismissed because even after a forensic blockchain investigation, none of the stolen funds have been traced to any named Defendant. ECF No. 15 at 31-32. He cites Oakdale Village Group v. Fong‘s holding that a plaintiff alleging conversion must “show an assumption of control or ownership
A preliminary peek at Brewer‘s arguments concerning specific causes of actions does not clearly show that those arguments are likely to succeed. Alternatively, it is not likely that leave to amend them to cure any deficiencies would be futile.
3. Conclusion
A preliminary peek at the motions to dismiss does not suggest that the arguments therein, individually or collectively, would be dispositive of this case without additional discovery. “Generally, there must be no question in the court‘s mind that the dispositive motion will prevail, and, therefore, discovery is a waste of effort.” Flynn v. Nevada, 345 F.R.D. 338, 346 (D. Nev. 2024) (emphasis in original). The predominant two-factor test does not support granting a stay.
C. SKELLERUP FACTORS AND THE BURDEN OF DISCOVERY
The Skellerup factors also do not favor staying discovery.6 Several of these factors are self-evident as applied to the relevant circumstances. As discussed above, the motions to dismiss are based on both questions of law and of fact. See supra III.B.2. The case is complex insofar as it involves three Defendants, a Plaintiff representing thirteen alleged victims, and seven causes of action based in both federal and state law. See supra III.B. Both of these factors then weigh against a stay. However, all three Defendants have moved for this stay, and none of them have
Defendants cite United States v. Dynamic Medical Systems, LLC and United States ex rel Modglin v. DJO Global Inc. to argue that the interest in avoiding unnecessary discovery is heightened when, as here, all claims sound in fraud and are subject to a heightened pleading standard. See supra III.B.1; ECF No. 44 at 10 (citing 2020 WL 3035219, at *5 (E.D. Cal. June 5, 2020); 2014 WL 12564275, at *2-3 (C.D. Cal. Feb. 20, 2014)). Neither case squarely applies here. Mogdlin concerned the specific context of False Claims Act (“FCA“) actions, which by definition involve “insiders privy to a fraud on the government” who as “insiders should have adequate knowledge of the wrongdoing at issue” to “be able to comply with Rule 9(b).” 2014 WL 12564275, at *2 (quoting Bly-Magee v. California, 236 F.3d 1014, 1019 (9th Cir. 2001)). As Modglin noted, under the FCA, “courts have repeatedly refused to allow qui tam relators to rely on later discovery to comply with Rule 9(b)‘s pleading requirements[.]” Id. (quoting United States ex rel. Karvelas v. Melrose-Wakefield Hospital, 360 F.3d 220, 231 (1st Cir. 2004)). “For these reasons, California district courts have found that it is proper in FCA cases where a motion to dismiss for failure to plead fraud with particularity is pending to stay discovery until the court has had the opportunity to decide whether the complaint satisfies the heightened pleading requirements of Rule 9(b).” Id. The logic of Modglin, turning as it does on the special insider knowledge qui tam relators are expected to have from the outset, does not extend to a case like this where the alleged fraud involved labyrinthine software.7
Nor does Dynamic Medical Systems suggest a special rule regarding stays of discovery whenever Rule 9(b) might apply. Rather, the judge in that case simply “decline[d] to potentially
Defendants then argue that because the District Judge vacated the scheduling conference pending resolution of this motion, Plaintiff cannot be prejudiced in the absence of a scheduling order with a discovery deadline. ECF No. 44 at 11 (citing ECF No. 37; Stavrianoudakis v. United States Dep‘t of Fish & Wildlife, 2019 WL 9667685, at *3 (E.D. Cal. Dec. 20, 2019)). While this is true, the court in Stavrianoudakis weighed this against the “unnecessary motion practice, litigation costs, and a waste of judicial resources” that would occur without a stay. 2019 WL 9667685, at *3. The question becomes what discovery Plaintiff seeks to propound before the District Judge may rule on the motions to dismiss.
Defendant argues that the discovery propounded on third-parties thus far demonstrates that the anticipated discovery would be burdensome and invasive. ECF No. 44 at 12. After St. George disclosed investments in real estate property used by an RV business, Plaintiff subpoenaed Rusty‘s, an RV business owned by Brewer‘s father. Id. (citing ECF Nos. 16-1 at ¶ 4, 39-4). This subpoena seeks a variety of records since May 2024, despite Rusty‘s apparently having no direct connection to this case. ECF No. 39-4 at 7, 10-11; ECF No. 44 at 12. Similarly, Plaintiff has subpoenaed Coinbase for information about eight cryptocurrency addresses, despite Brewer stipulating that one of them is his. ECF No. 39-3; ECF No. 44 at 13. Defendants categorize these subpoenas as “egregious example[s] of a fishing expedition and harassment“. ECF No. 44 at 14. They therefore request, at minimum, that discovery be limited to accounts, addresses, or businesses with no connection to currently named Defendants. Id.
Plaintiff argues that Defendants do not have “standing to quash a subpoena served upon a third party.” ECF No. 44 at 15 (quoting California Sportfishing Prot. All. v. Chico Scrap Metal, Inc., 299 F.R.D. 638, 643 (E.D. Cal. 2014)). While only the responding third-party can move to quash or for a protective order, this does not mean that their burden is irrelevant. While the undersigned has some concerns, based on the current record, about the subpoena to Rusty‘s, those
Plaintiff also argues that courts do not recognize the costs of responding to discovery as a “sufficiently particularized or unique harm to support a discovery stay where it would not otherwise be warranted.” ECF No. 44 at 21 (Elliot, 2024 WL 1676632, at *2). Defendants, however, do not just assert financial costs. They emphasize the pattern of fishing, purportedly harassment, that Plaintiff engages in through third-party subpoenas. ECF No. 44 at 14. This may constitute a more particularized form of harm than comprehensive discovery requests otherwise entail.
Defendants’ argument ultimately fails, however, because of the context in which Plaintiff propounded subpoenas for the information at issue. Brewer‘s motion to dismiss alleges that the FAC‘s failure to articulate what Defendants did with the stolen TAO is fatal to some of the claims. See supra III.B.2; ECF No. 15 at 31-32. Discovery regarding any businesses with a connection to named Defendants is therefore relevant, at a minimum for determining whether any investments are tied to the missing cryptocurrency. Defendants’ alternate proposal, allowing discovery into any company with no connection to them whatsoever, is unhelpful to such an inquiry.
Defendants have failed to demonstrate that the lack of a stay would sufficiently prejudice themselves or related entities. At the same time, the granting of a stay can be expected to delay the ultimate resolution of this case. The Skellerup Factors do not justify staying discovery.
IV. CONCLUSION
For the foregoing reasons, IT IS HEREBY ORDERED THAT:
- Defendants’ Motion to Stay Discovery in its entirety, or alternatively to stay any discovery not pertaining to the identification of Doe Defendants (ECF No. 39) is DENIED, except as to jurisdictional discovery.
- All third-parties who have withheld further responses to pending discovery on the outcome of Defendants’ Motion should resume production, or alternatively to move
to quash or for a protective order based on the merits of the specific discovery requests at issue.
IT IS SO ORDERED.
DATED: August 18, 2025
SEAN C. RIORDAN
UNITED STATES MAGISTRATE JUDGE