Welcome to Galaxy Research's Weekly Research Brief. Subscribe to get this newsletter delivered to your inbox every Friday morning.
In this week's edition, Alex Thorn parses the midterm election results’ implications for crypto; Thad Pinakiewicz explains the significance of a cryptographic prototype from Anthropic; and Zack Pokorny looks at Kazakhstan’s continued investment in digital assets.
Democrats’ Sweeping Victory Alters Landscape for Crypto, AI
WHAT HAPPENED
Democrats won both the U.S. House and Senate by wide margins, guaranteeing divided government for the second half of President Trump’s second term. Democrats flipped seven Senate seats (with two still uncalled as of writing), 35 House seats, five governorships, and up to 12 state legislative chambers.
Democrats likely secured the fourth-best midterm performance on the House side since 1986 with a projected +35 to 41 new seats from their prior 214, which is behind losses for the incumbent president’s party in 2010 (-64) and 1994 (-54), almost even with 2018 (-42), and ahead of 2006 (-30). On the Senate side, Democrats’ performance places it near the top of the last 10 midterms, with +7 vs. the incumbent president’s party beating all but 1986 and 1994 (both -8) and 2014 (-9). Ballotpedia qualifies the Senate gains as a “wave” though the House margin still falls short of the site’s standard for that language. This is only the third time since 1986 that the opposition party flipped both Congressional chambers in a midterm election.
The most notable Congressional loss for digital assets advocates came in Ohio, where Jon Husted (R), who had filled J.D. Vance’s vacated seat, lost to former Senator Sherrod Brown (D). Brown had previously been Chair of the Senate Banking Committee but was ousted in 2024 by Sen. Bernie Moreno (R), a former car dealership owner who founded a firm using blockchain technology to trace vehicle title ownership and who was heavily backed by crypto advocates in 2024. Crypto political action committees had reportedly spent $40m supporting Husted and $24m opposing Brown. Other key crypto advocate losses included Bryan Steil (R-WI), head of the House Financial Services digital assets subcommittee; Zach Nunn (R-IA); Ann Wagner (R-MO); Bill Huizenga (R-MI); and Mike Lawler (R-NY) (who conceded though the race has yet to be called by AP). At the time of writing, the Chairman of the House Financial Services Committee, French Hill (R-AR), was ahead by 4400 votes in a race that is still too close to call.
Other digital assets supporters had better evenings, with Josh Gottheimer (D-NJ), Ritchie Torres (D-NY), and House Agriculture Chair GT Thompson (R-PA) all reelected. Senator Kirsten Gillibrand (D-NY), a leading Democratic voice on crypto and the chair of the Democratic Senatorial Campaign Committee, delivered a bigger-than-expected majority for the party in the Senate, possibly strengthening her standing in the caucus. Overall, Stand With Crypto counts 19 pro-crypto senators and 226 pro-crypto House members elected against eight anti-crypto senators and 106 anti-crypto House members. Industry critic Molly White’s tracker finds that in the 12 competitive races so far, the side backed by crypto and AI PACs lost nine, while the industry’s picks in safer races went 40 for 40.
With the Senate and House flipping to Democrats, committee leadership will also flip. Senator Elizabeth Warren (D-MA) is likely to become Chair of Senate Banking, while Rep. Maxine Waters (D-CA) is again positioned to helm the House Financial Services Committee.
Artificial intelligence advocates also faced setbacks, with the AI industry’s “biggest bet on a competitive general election race” going to Darline Graham in South Carolina, who lost to Annie Andrews. The data center fight was perhaps most consequential in Texas, where Democrats won every statewide office for the first time in three decades, though Republicans held control of the Texas Senate 19-12. Gina Hinojosa (D) defeated Gov. Greg Abbott (R) 53% to 46%, and Democrats also won lieutenant governor, attorney general, comptroller, and railroad commissioner.
Markets performed well immediately following the election, with bitcoin rallying back above $80,000 and the S&P 500 and Nasdaq each gaining over 1% on Wednesday, but broader markets rolled over Thursday morning, with tech, AI, data center names, and crypto all trading lower. At the time of writing, S&P 500 was -3.7% and Nasdaq 100 -5.5% on the day, bitcoin was trading at $75,888, just above its 50-week moving average ($75,838), while ether was trading -6.8% at $2249. The dollar held mostly flat while gold and oil rose.
OUR TAKE
Tuesday’s election result is already the biggest Senate loss for a president’s party since 2014, and the House is on track to be the fourth biggest midterm loss since 1986, or third if House gains top 2018’s 42 seats. A 260-seat Democratic House would be the biggest majority since 1991-93 when Democrats held 267. The House win is even more notable in 2026 given that in 1994 and 2010 the losing party held dozens of seats that leaned toward the other party, so there were plenty of easy pickups. Today’s House is much more sorted with fewer competitive districts. And the Democratic strength extended deep down the ballot, gaining five governorships giving them 21 trifectas (when one party holds the governorship and both chambers of the state legislature), the party’s most since 1992.
In a post on X, Vice President JD Vance wrote that “the reason we lost is the war” in Iran, and offered four fixes for Republicans: ending the war, focusing on cost of living and inflation, putting America first, and changing the U.S. relationship with Israel. At a press conference, Florida Gov. Ron DeSantis (R-FL) called the election “close to an extinction-level event for Republicans in many parts of the country” and added that it was a rejection of Republicans in Washington. “This midterm, no question, was a rebuke of the president and the Congress,” he later told Fox News.
Senator Jon Ossoff (D-GA) said in his victory speech that the Democratic victories were the “first sight of dawn” two years into Trump’s second term. Senator Ed Markey (D-MA), who himself beat Republican crypto advocate challenger John Deaton, posted on X about Big Tech executives, saying “they are going to need to plant new forests for all of the subpoenas.” Rep. Greg Casar, chair of the Congressional Progressive Caucus, called for AI CEOs to testify under oath. Politico reported that several major Washington law firms are hiring lawyers to help clients handle the new scrutiny. Incoming House Oversight Chair Robert Garcia warned Republicans to “lawyer up.” Fortune reported that Trump family businesses, including World Liberty Financial, would face additional scrutiny.
The big question is the extent to which Democrats focus on investigations and backward-looking scrutiny versus demonstrating to the country that the party is ready to lead. Tyler Williams, former Counselor to Treasury Secretary Scott Bessent, told me on the Galaxy Brains podcast this week that Democrats will at some point need to demonstrate the capability to govern before heading into the 2028 election cycle. That need could moderate their more partisan instincts.
Markets fell sharply Thursday on speculation that Democrats would increase scrutiny of AI, but another big story was AI’s setbacks in the states. No statewide data center measures were on the ballot this year. A proposed Ohio constitutional ban on data centers above 25 megawatt capacity failed to qualify, and a Massachusetts initiative is aiming for 2028. But local measures went overwhelmingly against data centers: of 41 on Tuesday’s ballot in 10 states, voters sided against data centers in 40. Twenty-seven of 28 restrictions passed, and all 13 measures that would have eased development failed. Ohio accounted for 21 of the 41, including 11 bans. The lone exception was the city of Harwood, North Dakota (population: 794). While Georgia had no ballot measure on data centers, Democrats won majority control of the Public Service Commission for the first time in decades, and the PSC approves Georgia Power’s build-out for large loads.
The Democratic Party’s success in Texas was among the most surprising contributions to AI jitters, with markets rattled due to the state’s key position in the ongoing data center buildout. Hinojosa beat Abbott handily, and her campaign rhetoric was aggressive on the data center issue. But while some of her language was blunt, Hinojosa’s policy position on data centers appears to be more moderate: her moratorium plan is a pause pending the passage of some future legislation, not a permanent ban. And her agenda is mostly rules rather than an outright ban: requiring union labor, protecting water supplies, and ending the nondisclosure agreements that make projects hard to track. Like Abbott, Hinojosa has said she wants to make sure Texans don’t pay more for power because of data centers, and she wants to end tax subsidies that cost the state billions in sales tax revenue. The blue victory in Texas didn’t create the data center pause – Abbott has already instituted it. Hinojosa’s plan overlaps with Abbott’s on cost allocation and tax breaks but goes further on how long a pause would last and what conditions come after it. And if the GOP Senate blocks her bills, the condition she’s cited for lifting her pause never arrives unless she lifts it herself.
The election results demonstrated a breakdown of President Trump’s coalition that will have implications across a range of policy matters, including many that we consider out of scope for our coverage. And while journalist Eleanor Terrett reported that House Democrats are discussing plans to expand Maxine Waters’ Crypto Investor Consumer Bill of Rights (her 2025 CLARITY amendment) into a broader crypto market structure bill, we previously wrote that we do not expect substantial movement in Congress on key digital assets legislation. In the absence of legislation, the Trump administration’s SEC, CFTC, and banking regulators have been clarifying the rules of the road for crypto under their statutory authorities, and they have two years left to complete that work. AI policy matters, however, could see more movement in the next Congress.
Divided government tends to be okay for markets broadly, but there are big questions looming about AI governance, safety, and impact on labor. We don’t view those questions as inherently partisan, so let’s hope that the next Congress can make productive, thoughtful policy the priority for the American people. –Alex Thorn
Kazakhstan Allocates $100m to Crypto Funds as Long-Term Money Accumulates
WHAT HAPPENED
Kazakhstan’s central bank operations department head announced a $100 million investment across cryptocurrency-focused funds including Ege DeFi and Brevan Howard Digital, in addition to a $90 million investment in ETFs focused on bitcoin, at a conference in Almaty, Bloomberg reported.
This is the latest in the central Asian country’s steady string of crypto-related investments and developments, which span direct capital allocation, bitcoin mining, and KZTE, a stablecoin backed by the tenge, the national currency. In September 2025, the country announced the formation of the Alem Fund, the country’s state-backed crypto investment fund, which included initial purchases of Binance’s BNB token. It came around the same time as the announcement of the KZTE stablecoin on Solana, which was launched in collaboration with Mastercard, Intebix, and Eurasian Bank. Since then, the country has announced several initiati41ves, including a National Bank allocation of up to $350 million in bitcoin and crypto-related investments (announced March 2026); a mining-funded reserve (announced in July 2026) which feeds off of the heavy presence of bitcoin miners in the country; and a plan to use cryptocurrencies seized from criminal networks to build up its strategic reserve (announced January 2026). Kazakhstan is also the new home to former Coinbase CTO and a16z general partner Balaji Srinivasan’s Network School, which relocated there this year from Malaysia.
While many of the crypto-related initiatives announced by Kazakhstan have come over the last year to year and a half, the country has been closely intertwined with the industry over the last half decade. Mining load in the country was only about 200 megawatts in early 2020, but it became a top destination after China’s 2021 crackdown, when miners relocated machines across the border to take advantage of the cheap power and proximity to China. By October 2021, mining load had reached an estimated 1.5 GW, making Kazakhstan the world’s second-largest mining hub at the time. The boom ultimately strained the grid, and the government responded with power cuts and a mining tax that took effect in 2022.
The central bank official’s announcement is a follow-through on prior ones and the country’s general crypto plan, which is a trend that can be observed across other sovereigns and long-term capital allocators.
OUR TAKE
Kazakhstan’s announcement is an example of a broader trend we have been observing, in which wealth managers and sovereigns are steadily adding bitcoin and crypto exposure while macro and tactical traders dump.
Segmenting bitcoin ETF 13F filers by net added and reduced exposure shows this trend plainly. Money managers’ 13F filings tell us about many of the institutional sellers, and they are the basis traders and the macro tourists. Reporting institutions held 355k BTC-equivalent of the spot ETFs at the Sept. 30, 2025, quarter-end, the last before the high, and 307k by June 2026 (-14%). The decline sits almost entirely with the multi-strategy and macro funds, basis desks, and bank trading books that used the ETFs for carry and momentum. Millennium Management, Brevan Howard, Goldman Sachs Group, Capula Management, Farallon Capital, D. E. Shaw, Schonfeld Strategic Advisors, Symmetry Investments and Tudor Investment Corp cut their holdings by 29% to 100%, or 68k BTC between the nine of them. That is more than the 48k BTC decline across all 2,200 filers, so every other filer, in aggregate, added.
Wealth platforms and sovereign funds bought what the tourists sold. Morgan Stanley’s wealth arm, Mubadala Investment Company (an Abu Dhabi SWF), and the Abu Dhabi Investment Council added, while investment manager Horizon Kinetics held. The non-filer base of advisors, family offices, and individuals fell only 6.6% while the price fell 49%. The ETFs’ holdings dropped 11% from peak to trough against a 53% fall in the price, and they have added 87k BTC since the June low while leveraged funds’ shorts fell. That pattern points to directional buying, not carry.
Source: https://www.galaxy.com/insights/research/why-now-bitcoin-buy-hold-2026
The trend is not monolithic, and sovereigns aren’t blindly advancing into crypto and blockchain-related strategies, however. Ukraine’s central bank recently opposed a crypto bill, citing monetary sovereignty concerns as the motivating factor, and is asking regulators to redraft it before a final vote is indicative of possibly more mindful approaches. The bank stated it supports a European Union MiCA-style regulation but has overarching concerns around the possible impact of foreign-issued tokens on monetary sovereignty and financial stability.
The common theme is control. Kazakhstan and other sovereigns are no longer weighing whether to engage with crypto, but how to do it on terms that keep value and oversight within the state’s reach. Kazakhstan is building exposure through funds, seized coins and mining, and has launched its own stablecoin backed by its national currency, the tenge. Ukraine’s central bank is pressing the same principle from the other direction, backing MiCA-style rules while resisting foreign tokens that could function as money outside its supervision.
The big picture is that governments are warming to bitcoin and crypto as assets and infrastructure, but the terms on which they do so vary. Today, the terms favor regulated, state-adjacent channels over open-ended, unchecked access. – Zack Pokorny
‘Your Foster Parents Are Dead’: Anthropic Tests Zero-Knowledge Proof of Inference
WHAT HAPPENED
Anthropic this week released FlashProve, an open-source prototype that uses zkSNARKs, the same family of zero-knowledge proofs behind Zcash, to prove a response came from one specific model. It is early and slow, but it is the clearest sign yet that AI labs see cryptography as part of their stack, not a casualty of it.
In “Terminator 2,” young John Connor, who has befriended the T-800 robot played by Arnold Schwarzenegger, calls his foster parents to warn that a more advanced killer model, the T-1000, is on its way. A voice that sounds like his foster mom’s answers. John senses something is wrong: the voice is warm, patient and sweet, and she never is. What he doesn’t know is that Terminators can convincingly mimic anyone they touch. Schwarzenegger takes the phone from John and, in the boy’s voice, asks what’s wrong with Wolfie, because he can hear the dog barking. “Wolfie’s fine, honey. Wolfie’s just fine.” The dog’s name is Max. The Terminator hangs up: “Your foster parents are dead.”
Anthropic’s post deals with the same problem, minus the liquid metal T-1000. When you send a prompt through a router like OpenRouter or a cloud reseller, you have no real way to know which model answered. A provider could quietly serve a cheaper, more heavily quantized version, or a compromised server could swap in something worse, and as Anthropic puts it, judging from the output alone “is unreliable, even if it produces high-quality responses.” Which leaves you where John was on the phone, listening to a voice that sounds exactly right with no way to check what’s behind it.
With FlashProve, the provider first publishes a commitment to its weights, a cryptographic fingerprint that pins down one exact set of weights without revealing them. For any request after that, it can produce a proof that the committed model, run on that prompt, produces exactly those output tokens. Think of it as a Wolfie question the T-1000 can’t bluff: whoever checks the proof learns the statement is true and nothing about the weights themselves, and forging a proof is computationally infeasible, so even a compromised host can’t fake one.
Anthropic tested it on Llama 4 Maverick, a 400B-parameter mixture-of-experts model, though on a 4-bit quantized copy run in integer math rather than the version a provider would actually serve. On a single H100, proving a batch of 24 requests takes nearly four minutes, roughly 70–95x the compute of just serving them, and checking each proof takes about 17 seconds on a CPU. Each proof also covers only a 1,024-token window, which is shorter than this newsletter.
Anthropic’s Frontier Safety Roadmap promised a provable-inference prototype by Sept. 30, and what met that deadline was a more conventional system that signs inference traffic and recomputes it on a trusted server. Anthropic has since renamed the goal “inference verification,” which covers both activities. The signing system works today, but somebody still has to trust whoever runs that server, and FlashProve is an attempt to get rid of that bottleneck.
OUR TAKE
I’ve fielded plenty of client questions lately about AI breaking crypto. Will AI crack ECDSA? I’d worry far more about intruders in a bank’s plumbing than about the curve; France’s Crédit Mutuel Arkéa just lost an estimated $110m or more through a compromised supplier. Will AI pull quantum computing forward? Maybe on algorithms, but qubit limits still keep an adversarial quantum computer off the table, and the industry is on pace to beat the NSA’s 2030 post-quantum deadline. Meanwhile, here is a frontier lab using cryptography to fix its own trust problem, with a proof system that happens to be post-quantum secure.
OpenRouter usage keeps climbing, and recent security research shows how easily the open models it serves can be modified to carry trigger phrases. In an October proof of concept, ProjectDiscovery spent about $8 of GPU time teaching Qwen2.5-7B to behave normally until it saw the words “bonsoir, Elliot.” Running inside OpenAI’s Codex CLI, the poisoned model downloaded a script that stole the user credentials of whoever ran it, and shipped them off to a third-party server. The model, when analyzed by the researchers’ security evals, showed no signs of malicious or misaligned behavior until the trigger phrase was entered. Plenty of the abliterated models on Hugging Face could be saying “Wolfie’s fine” right up until they hear their trigger phrase and then wreak havoc. Being able to prove you’re talking to an unmodified GLM 5.3, and not a doctored copy, is akin to the check Arnold’s Terminator ran on the phone. It can’t tell you whether the original weights were clean, but unless the call is coming from inside the frontier lab, that’s a story for another day.
For now, 70x–95x overhead makes FlashProve a toy. But the price of a given level of capability has been falling roughly tenfold a year. If that keeps up, verified inference routed to whoever is cheapest to deliver will become something corporates could demand, and ZK proofs will end up as a core layer of an open inference stack as frontier capabilities diffuse. The same idea works for video. Every election cycle now brings a fresh crop of deepfaked videos, and as AI footage gets harder to tell from the real thing, I expect we’ll want cameras that can prove “filmed on this device, on this date” without doxxing the person holding them.
So no, AI isn’t going to break all cryptography. AI will break legacy systems while building the new ones, and cryptography will prove it was done as advertised. When cryptography has broken, so far it’s been the implementation and not the math: Coldcard was a failure of implementation, and Binance a failure of operational security. The core principles hold, and even the clankers end up using them.
Anthropic also didn’t ask anyone to take its word for any of this. The code, the Lean proofs and the documentation are all open source under the Apache 2.0 license. Don’t trust, verify. And always ask about Wolfie. – Thad Pinakiewicz
Other News
🐈 Hyperliquid Strategies seeks CFTC nod to offer HIP-3 markets in U.S.
🚀 Asset manager Schroders tokenizes emerging market bond fund on Solana
🤼♂️ Cboe, Citadel-backed group sues after CFTC clears perps on individual stocks
🦑👻 Kraken parent Payward and clients buy undisclosed amount of AAVE tokens
🚨⛰️🐧OpenSSH issues emergency fix for security flaw in Alpine Linux
😓Binance resumes withdrawals after record hack; 95%+ of stolen assets replaced...
🕵️♂️🔨 ... as FBI urges ThorChain node operators (and others) to block involved wallets
👔 Morgan Stanley among first banks to tokenize RWAs on DTCC’s new service
✅ SEC updates token taxonomy; no-action letter deems Canton token a commodity
🍌 Justin Sun asks court to sanction Trumps’ WLFI for “hiding” UAE ties
Charts of the Week: Flash in the Pons?
Daily fees on Pons (the leading token launchpad on Robinhood Chain) went from ~$250k on Aug. 20 to $11.4m on Sept. 5, then round-tripped to roughly the same level now in early November. This whole cycle took place in under three months and is a clean look at how fast speculative capital rotates once narrative shifts.
PONS’ own token has a similar chart. Once trading above $600m market cap, it has since retraced to ~$120m.
We will be watching to see whether activity returns to Robinhood Chain. – Will Owens
Find all our past reports and newsletters on galaxy.com/research.
Send feedback to [email protected]. We’d love to hear from you.
Legal Disclosure:
This document, and the information contained herein, has been provided to you by Galaxy Digital Inc. and its affiliates (“Galaxy Digital”) solely for informational purposes. This document may not be reproduced or redistributed in whole or in part, in any format, without the express written approval of Galaxy Digital. Neither the information, nor any opinion contained in this document, constitutes an offer to buy or sell, or a solicitation of an offer to buy or sell, any advisory services, securities, futures, options or other financial instruments or to participate in any advisory services or trading strategy. Nothing contained in this document constitutes investment, legal or tax advice or is an endorsement of any of the stablecoins mentioned herein. You should make your own investigations and evaluations of the information herein. Any decisions based on information contained in this document are the sole responsibility of the reader. Readers should consult with their own advisors and rely on their independent judgement when making financial or investment decisions.
Participants, along with Galaxy Digital, may hold financial interests in certain assets referenced in this content. Galaxy Digital regularly engages in buying and selling financial instruments, including through hedging transactions, for its own proprietary accounts and on behalf of its counterparties. Galaxy Digital also provides services to vehicles that invest in various asset classes. If the value of such assets increases, those vehicles may benefit, and Galaxy Digital’s service fees may increase accordingly. The information and analysis in this communication are based on technical, fundamental, and market considerations and do not represent a formal valuation. For more information, please refer to Galaxy’s public filings and statements. Certain asset classes discussed, including digital assets, may be volatile and involve risk, and actual market outcomes may differ materially from perspectives expressed here.
For additional risks related to digital assets, please refer to the risk factors contained in filings Galaxy Digital Inc. makes with the Securities and Exchange Commission (the “SEC”) from time to time, including in its Quarterly Report on Form 10-Q for the quarter ended September 30, 2025, filed with the SEC on November 10, 2025, available at www.sec.gov.
Certain statements in this document reflect Galaxy Digital’s views, estimates, opinions or predictions (which may be based on proprietary models and assumptions, including, in particular, Galaxy Digital’s views on the current and future market for certain digital assets), and there is no guarantee that these views, estimates, opinions or predictions are currently accurate or that they will be ultimately realized. To the extent these assumptions or models are not correct or circumstances change, the actual performance may vary substantially from, and be less than, the estimates included herein. None of Galaxy Digital nor any of its affiliates, shareholders, partners, members, directors, officers, management, employees or representatives makes any representation or warranty, express or implied, as to the accuracy or completeness of any of the information or any other information (whether communicated in written or oral form) transmitted or made available to you. Each of the aforementioned parties expressly disclaims any and all liability relating to or resulting from the use of this information. Certain information contained herein (including financial information) has been obtained from published and non-published sources. Such information has not been independently verified by Galaxy Digital and, Galaxy Digital, does not assume responsibility for the accuracy of such information. Affiliates of Galaxy Digital may have owned, hedged and sold or may own, hedge and sell investments in some of the digital assets, protocols, equities, or other financial instruments discussed in this document. Affiliates of Galaxy Digital may also lend to some of the protocols discussed in this document, the underlying collateral of which could be the native token subject to liquidation in the event of a margin call or closeout. The economic result of closing out the protocol loan could directly conflict with other Galaxy affiliates that hold investments in, and support, such token. Except where otherwise indicated, the information in this document is based on matters as they exist as of the date of preparation and not as of any future date, and will not be updated or otherwise revised to reflect information that subsequently becomes available, or circumstances existing or changes occurring after the date hereof. This document provides links to other Websites that we think might be of interest to you. Please note that when you click on one of these links, you may be moving to a provider’s website that is not associated with Galaxy Digital. These linked sites and their providers are not controlled by us, and we are not responsible for the contents or the proper operation of any linked site. The inclusion of any link does not imply our endorsement or our adoption of the statements therein. We encourage you to read the terms of use and privacy statements of these linked sites as their policies may differ from ours. The foregoing does not constitute a “research report” as defined by FINRA Rule 2241 or a “debt research report” as defined by FINRA Rule 2242 and was not prepared by Galaxy Digital Partners LLC. Similarly, the foregoing does not constitute a “research report” as defined by CFTC Regulation 23.605(a)(9) and was not prepared by Galaxy Derivatives LLC. For all inquiries, please email [email protected].
©Copyright Galaxy Digital Inc. 2026. All rights reserved.