Crypto News

Crypto trader Machi Big Brother (@machibigbrother) extended his winning streak from 10 to 12 consecutive profitable trades over five days to one week, according to blockchain analytics platform Lookonchain. His cumulative gains rose from about $1.34 million to approximately $2.14 million, with two additional profitable PUMP trades accounting for the latest increase. Machi Big Brother’s current long positions have also expanded to more than $150 million. They include about 34,100 ETH worth $92.98 million, 456 BTC valued at $39.41 million, 174,500 HYPE worth $15.84 million and 425 million PUMP valued at $2.72 million. The earlier reported positions were approximately 33,950 ETH, 409 BTC and 180,000 HYPE, worth about $144 million in total. The trader’s large crypto holdings highlight the potential risks of concentrated and leveraged positions. However, Lookonchain’s data does not confirm Machi Big Brother’s overall realised or unrealised profit, leverage or liquidation risk. There is also no clear evidence that the trades have materially affected ETH, BTC, HYPE or PUMP prices, or that the trader has changed his broader strategy. Traders should monitor position reductions, additions and liquidations for possible short-term volatility signals.

Former Anthropic researcher Jacob Coxon is scheduled to testify before the New York City Council on Monday, 5 October 2026, at a hearing focused on AI safety and regulation. Representatives overseeing policy and safety from Anthropic, OpenAI, Google and Meta are also expected to attend, alongside former Google DeepMind researcher Alex Turner and Daniel Koktailo of the AI Futures Project. Coxon left Anthropic in September and warned that advanced AI could become too powerful for humans to control. He accused Anthropic and OpenAI of “betting our lives” in the race to build more capable systems. His view contrasts with OpenAI co-founder Sam Altman, who has rejected claims that AI presents an “extremely high” risk of catastrophe. The hearing, convened by New York City Council Speaker Julie Menin as a Committee of the Whole, will involve all 51 council members. Lawmakers are considering measures including third-party validation of AI systems used by the city, stronger whistleblower protections and incentives, and a private right to sue over AI-related harm. The session could become the first time a US legislative body compels sworn testimony on AI safety. The event may influence future AI governance, model-safety standards, corporate accountability and technology policy. For crypto traders, it is primarily a regulatory and sentiment signal rather than a direct token catalyst. Short-term price effects on AI-related crypto projects are likely to be limited, although stricter rules could weigh on AI-linked tokens and infrastructure providers if regulatory concerns intensify.

US President Donald Trump has appointed National Intelligence Director Jay Clayton to lead the new Super Intelligence Force, according to a post on Truth Social. Clayton will report directly to Trump and White House Chief of Staff Susie Wiles while potentially retaining his intelligence role. The Super Intelligence Force will coordinate federal artificial intelligence and superintelligence initiatives, following a model similar to the US Space Force. Trump also appointed Federal Trade Commission Chair Andrew Ferguson, Office of Personnel Management Director Scott Kupor and Defence Department technology chief Emil Michael as deputies. The appointments indicate that the Super Intelligence Force may cover AI regulation, government staffing, defence research and technology policy. The initiative could support spending on AI infrastructure, data centres, computing power and federal technology contracts. It may also create longer-term opportunities for blockchain analytics, algorithmic trading and AI-agent services. No specific budget, procurement programme or cryptocurrency policy has been announced. The immediate impact on major cryptocurrencies is therefore likely to remain limited. Traders should monitor federal AI budgets, contract awards and regulatory standards, which could influence AI-related equities and tokens over time.

Hyperliquid’s HYPE buyback and burn programme accelerated. The platform initially bought and permanently burned 20,200 HYPE at a volume-weighted average price of $89.05, worth about $1.8 million. In the latest update, Hyperliquid burned 112,580 HYPE at an average price of $90.20, valued at approximately $10.15 million. The latest tokens were reportedly funded by revenue and fees generated through AQAv2. Cumulative HYPE burns have reached about 49.25 million tokens, worth an estimated $4.45 billion and representing roughly 4.93% of the token’s maximum supply. Hyperliquid generated about $989,000 in estimated fees over 24 hours, $9.85 million in revenue over seven days and $52.06 million over 30 days. The HYPE burn reduces the available token supply and signals continued fee generation on the Hyperliquid trading platform. This may support HYPE over the longer term, but traders should also monitor trading volumes, investor sentiment, token unlocks and broader crypto-market conditions, which could offset the deflationary effect.

The US midterm elections on 3 November could reshape crypto regulation, congressional oversight and tax policy. Polling aggregator 270toWin, Kalshi and Polymarket indicate that Democrats are favored to regain the House, while control of the Senate remains uncertain. The result will affect oversight of the SEC, CFTC, OCC and Treasury, agency budgets and future crypto legislation. The stalled Clarity Act could receive renewed attention during a lame-duck session if Republicans retain the House. A bipartisan crypto tax bill has advanced through the House Ways and Means Committee, and Senator Steve Daines has introduced a Senate version. Crypto taxation is therefore likely to remain a major issue regardless of which party controls Congress. A Democratic-controlled House could increase scrutiny of crypto companies linked to Donald Trump and his business interests. Investigations, subpoenas and hearings could create reputational and compliance risks. Crypto-focused groups have committed at least $33 million to the election, including $30 million from Fairshake and $3 million from the Digital Freedom Fund. For crypto traders, the US midterm elections are a regulatory catalyst rather than an immediate price trigger. Republican control and progress on the Clarity Act could improve regulatory certainty. Democratic gains could bring tougher oversight, tax-policy risks and delays to legislation. The outcome may influence compliance costs, institutional participation and crypto capital flows, while short-term market reactions are likely to remain headline-driven.

Hyperliquid whale data shows a strong short bias among large traders. As of 5 October at 08:19 UTC+8, about 200 active accounts with positions above $3 million held roughly $830 million in BTC shorts, compared with $518 million in longs. The BTC long-to-short ratio was 0.62. The same accounts held about $1.05 billion in ETH shorts and $687 million in longs, giving ETH a 0.65 long-to-short ratio. Short exposure exceeded long exposure by more than 1.5 times in both markets. This Hyperliquid whale positioning points to a defensive outlook for BTC and ETH and could increase short-term volatility. A price rebound could trigger short liquidations and a squeeze, while weakening support may accelerate leveraged selling. The data covers a limited group of accounts and does not confirm the broader crypto market trend. Traders should also monitor funding rates, open interest, liquidation levels and spot-market flows.

SEC Chair Paul Atkins said the agency will propose more crypto regulation and digital-asset rules as the United States seeks to remain a global cryptocurrency hub. The SEC is reviewing custody requirements created in 1940 for traditional assets, arguing that current infrastructure can lag behind newly launched digital assets. The custody plan is part of a broader SEC crypto regulation programme developed since 2025. Other proposals include Regulation Crypto Assets, rules for tokenised securities and an Innovation Exemption. The proposed five-year sandbox could allow US stocks to trade on decentralised exchanges and through liquidity providers. The latest push follows the US Senate’s failure to pass the CLARITY Act. Former SEC internet-enforcement chief John Reed Stark said the agency could be exceeding its authority and bypassing Congress. For crypto traders, the SEC crypto regulation agenda may improve long-term legal clarity and institutional confidence, but it could create short-term volatility and higher compliance costs for issuers, custodians and trading platforms.

DeepSeek and Huawei have open-sourced six software modules for Huawei Ascend AI chips, targeting Nvidia’s CUDA advantage. The toolkit includes FlashMLA, DeepEP, DeepGEMM, TileKernels, DeepSelect and an Ascend version of TileLang. TileLang is the central component, offering a higher-level framework intended to simplify AI development while retaining hardware performance. The tools support many operators used to train DeepSeek’s V4 models. DeepGEMM supports BF16, FP8 and FP4 matrix operations, while TileKernels aims to provide compatible Python interfaces for Nvidia GPUs and Huawei NPUs. DeepSeek and Huawei have also optimised computing and communications for a 128-chip Ascend 950 cluster. The partnership is expanding. DeepSeek is preparing to deploy at least 160,000 Ascend 950DT chips at a data centre in Inner Mongolia. The chips are expected to launch in the fourth quarter of 2026, while parts of the facility could come online from late 2027 to early 2028. DeepSeek has already optimised its V4 model for Ascend hardware and used Ascend 950 chips in some training workloads. The release could reduce Chinese developers’ reliance on Nvidia and limit exposure to US export controls. However, the performance evidence remains limited. Early tests used an unpublished proof-of-concept development kit and manual configurations. Full bandwidth is expected only with Huawei’s commercial Atlas 850E kit, reportedly planned for October 2026. DeepEP currently reaches about 90% to 95% of physical bandwidth at expert-parallelism scales of up to 32, while larger configurations and data-combine operations are still being optimised. FlashMLA is not yet supported on Ascend, and DeepSelect currently supports only BF16. For crypto traders, this is a neutral and mainly long-term technology story rather than an immediate market catalyst. The DeepSeek partnership may support demand for domestic AI infrastructure, semiconductor alternatives and data-centre capacity, but it does not directly affect major cryptocurrency fundamentals. Traders should monitor TileLang adoption, Ascend shipments, DeepSeek’s future hardware purchases and Nvidia’s China market share.

Solana ETF inflows reached a weekly record of about $188 million through late September, with all seven US spot Solana ETF products reporting net inflows. Bitwise’s BSOL led demand with roughly $128 million, or 68% of the total. By October 4, SOL had recovered above $120 and traded at $120.96. Technical momentum remained constructive: the 14-day RSI stood at 65.4, SOL was above its key moving averages, and MACD showed a Buy signal. However, the latest advance lacked strong volume confirmation. SOL faces initial resistance at $122.57, followed by a key $123.35-$125 breakout zone. A sustained daily close above $125 could target $127.50-$128.00. Support is near $119.84, with further downside levels at $116.50-$116.90 and $112.54-$113.00. Earlier analysis placed support near $118 and highlighted mixed MACD readings, while Solana developers continued testing the Alpenglow upgrade, which aims to reduce transaction finality from about 12.8 seconds to roughly 150 milliseconds. Strong Solana ETF inflows improve the demand outlook, but traders may wait for higher volume before confirming a durable breakout.

Ethereum price prediction remains focused on whether ETH can reclaim $2,700 and secure a daily close above the $2,775–$2,800 resistance zone. ETH closed near $2,687 on October 4 after briefly reaching $2,765, leaving the market at a key technical decision point. A sustained breakout above $2,800 could target $2,900, $3,000 and potentially $3,050. Failure to hold $2,700 may expose support around $2,600–$2,650, while the broader support zone remains near $2,400–$2,500. A break below $2,560 would weaken the short-term bullish setup. Ethereum price prediction is also being shaped by mixed momentum indicators and weaker institutional demand. Monthly RSI at 51.53 shows moderate improving momentum, but MACD remains negative without a confirmed bullish crossover. US spot Ethereum ETFs recorded three consecutive sessions of net outflows through October 2: $59.6 million on September 30, $55.4 million on October 1 and $17.3 million on October 2. This followed about $118 million in outflows from September 28 to October 2, after roughly $689.8 million of inflows in the previous week. Earlier, BlackRock-linked funds bought about $1.57 billion worth of ETH over 20 days, highlighting the mixed institutional picture. The Glamsterdam upgrade is scheduled to activate on the Sepolia testnet on October 6. It targets Layer-1 scaling, block construction and more efficient gas pricing, with mainnet activation expected later in the fourth quarter. The upgrade could support longer-term sentiment, but traders are likely to prioritise the $2,800 breakout, ETF flows and support near $2,560 when assessing Ethereum’s next major move.

Ethereum (ETH) first moved above 2,700 USDT on OKX, reaching 2,700.19 USDT with a 0.49% 24-hour gain. Later market data on 4 October 2026 showed ETH at 2,700.21 USDT, with its 24-hour increase widening to 0.62%. The move keeps Ethereum above a key psychological price level and could attract short-term momentum traders. However, the limited gain does not yet confirm a decisive ETH price breakout or strong buying pressure. Traders should watch trading volume, broader crypto-market sentiment and whether Ethereum can hold above 2,700 USDT. Sustained support could encourage further upside, while a break back below the level may signal a failed breakout and renewed consolidation.

Brazil’s election has placed crypto regulation and compliance under scrutiny. In the final Datafolha poll before the October 4 first round, Luiz Inácio Lula da Silva led Flávio Bolsonaro 45% to 42% among valid votes. A runoff on October 25 appeared likely, with Lula narrowly ahead 47% to 46% in the latest runoff poll. Existing crypto rules will remain in force regardless of the result. Central Bank regulations require virtual-asset service providers to obtain authorization and meet governance, anti-money-laundering and foreign-exchange standards. From November 6, regulated institutions will generally be barred from dealing with unauthorized crypto firms. Financial institutions must also report covered transfers of at least $10,000 involving self-custody wallets to Coaf. Further rules on custody, reserves, staking and overseas transfers are planned for 2027. Lula’s re-election would likely extend the current crypto regulation path, while a Bolsonaro victory could increase policy uncertainty without guaranteeing deregulation. Flávio Bolsonaro’s platform contains no specific proposals on crypto, stablecoins or virtual assets. A postponed consultation on stablecoin taxation leaves future tax policy unresolved. Brazil was ranked first globally for grassroots crypto adoption in 2026, with estimated activity of $252.5 billion. For traders, the Brazil election is mainly a regulatory and compliance event rather than an immediate crypto price catalyst. Election headlines could briefly affect local exchange activity and stablecoin flows, while longer-term market effects will depend on licensing costs, enforcement and tax policy.

A wallet linked by Lookonchain to Andrew Tate transferred 20,950 HYPE, worth about $1.87 million, to Binance on 2 October. Earlier reports placed the transfer on 3 October, while the latest data dates it to 2 October and provides updated holdings. The wallet bought 122,826.9 HYPE at an average price of $4.48 on 29 November 2024, investing about $550,000. Its realised and unrealised HYPE profit is estimated at $7.24 million, a 1,317% return. The address now holds 42,417 HYPE, valued at roughly $3.7 million. The HYPE deposit could create short-term selling pressure if the tokens are sold, but the transfer does not confirm a sale or signal a broader market trend. Traders should monitor HYPE exchange inflows, trading volume and further wallet activity.

Anthropic is investing $100 million in the Claude Frontier Academy to train 10,000 enterprise AI deployment engineers by the end of 2027. The Claude Frontier Academy will train employees from customers and consulting firms, rather than Anthropic staff, to identify use cases, build AI systems, complete security reviews and train internal teams. Its first programme, the Frontier Deployed Engineer Residency, follows a medical-residency model. Participants receive practical training, lead a real Claude project within their organisation for 12 weeks and complete assessments before certification. Initial cohorts are operating in San Francisco, New York and London, with participants nominated by Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk. The first certified engineers are expected in early 2027. Accenture separately plans to train about 30,000 professionals on Claude. The Claude Frontier Academy targets a shortage of enterprise AI deployment talent, which Anthropic sees as a bigger barrier than model access. Its selective approach represents an average investment of about $10,000 per trainee. OpenAI plans to support 300,000 certified AI consultants through a $150 million partner programme, while Microsoft is assigning 6,000 employees to its Frontier Company. Embedding Claude-trained engineers and projects inside companies could improve enterprise adoption, demand for implementation services and customer retention. However, major consulting firms work with multiple AI providers, limiting Anthropic’s competitive moat. The programme has strategic significance for the AI sector but no direct cryptocurrency catalyst. Crypto traders should therefore view the news as neutral for digital-asset prices and monitor broader technology-sector sentiment instead.

Former UK National Crime Agency officer Paul Chowles has been ordered to forfeit £1,810,678.93 after stealing 50 Bitcoin from evidence seized during the Silk Road 2.0 investigation. The Bitcoin was worth about £59,409 in 2017, but its later rise increased the assessed criminal benefit to £1,970,759.27. Chowles transferred the Bitcoin, split the funds and used the Bitcoin Fog mixer to conceal the transactions. He also converted part of the holdings through Cryptopay and Wirex cards. Police linked the transfers to his iPhone, notes and account records after arresting him in 2022. He pleaded guilty in May 2025 and received a five-year-and-six-month prison sentence. Authorities recovered 30 of the 50 stolen Bitcoin. A separate confiscation order was issued against darknet operator Thomas White after the theft was initially attributed to him. The case highlights Bitcoin custody, blockchain tracing and the legal risks of mishandling seized cryptocurrency. It is unlikely to create direct selling pressure on Bitcoin, making the market impact neutral.

Grayscale’s Zcash ETF, ZCSH, recorded $93.56 million in net outflows in the week ended October 2, its first negative week since late August. Redemptions reached $30.25 million on September 30 and $26.93 million on October 2. The Zcash ETF’s cumulative net inflows fell to about $212.56 million, while total net assets declined to roughly $751 million from a peak near $915 million–$979 million. ZEC closed at about $1,304 on October 3, down around 21% from its September 26 close near $1,653 and below its recent high above $1,650. The outflows coincided with falling ZEC perpetual open interest, profit-taking and weaker spot momentum. However, the available data does not establish that Zcash ETF redemptions caused the entire decline. Continued withdrawals could still add selling pressure if the fund liquidates holdings. Grayscale converted its Zcash Trust into the ZCSH spot ETF on August 25. The fund completed a three-for-one share split on September 30, which reduced the per-share value without changing investors’ overall holdings. Digital Currency Group invested about $100 million through an authorised participant transaction on September 8. Traders should also monitor Zcash’s NU7 upgrade: testnet activation was expected around October 6, with mainnet targeted for November 5. The upgrade is designed to reduce target block times from 75 seconds to 25 seconds and revise fee sustainability and shielded-transaction protections.

Cronos Network said its community approved two CRO tokenomics proposals. The Cronos community pool has burned 228 million CRO, taking total CRO burns to 428 million. Under the updated CRO tokenomics plan, 100% of revenue from Cronos Ult and Cronos Launch will fund open-market CRO buybacks. The repurchased tokens will be burned monthly, and Cronos will publish transaction hashes for transparency. CRO staking rewards will not change and will remain funded by the Strategic Reserve. The buyback-and-burn programme could reduce circulating supply and create recurring demand for CRO over time. Traders should monitor ecosystem revenue, buyback size, burn data and broader market sentiment. The short-term price effect may remain limited if purchases are small or market conditions weaken.

Robert Kiyosaki says Bitcoin, gold and silver can serve as financial insurance against inflation, currency debasement and economic instability. The Rich Dad Poor Dad author describes himself as a financial prepper, stressing that preparing for a crisis does not mean predicting one. Kiyosaki prefers assets that governments cannot print. He highlights Bitcoin’s fixed supply of 21 million coins, while acknowledging that scarcity does not guarantee stable purchasing power. He has also warned about government debt, geopolitical risks and weaknesses in traditional retirement systems. Kiyosaki previously said he was accumulating BTC and ETH during market weakness, although his buying decisions have changed during earlier Bitcoin and Ethereum declines. For crypto traders, the comments support Bitcoin’s long-term inflation-hedge and alternative-store-of-value narrative. However, they provide no new market data or policy development, so the immediate price impact on Bitcoin is likely to be limited.

The Zano exploit exposed a Gateway Address validation flaw introduced in Hard Fork 6. The attacker paid only 100 ZANO, worth about $553 at the time, to register a gateway address and mint roughly 36.9 million counterfeit ZANO over two events on 29 August and 25 September. The attacker also created about 1.8 trillion fUSD stablecoins. The counterfeit assets behaved like legitimate tokens and could be transferred or traded normally. Zano said the incident potentially affected 117,941 outputs and 65,301 transactions. The project rolled back about one month of blockchain history from block 3,833,000, disabling Gateway Addresses and removing the illicit supply. It said the total ZANO supply and issuance schedule would remain unchanged. The rollback also reverses legitimate transactions made during the affected period. Zano plans to restore user balances using developer funds, team members’ personal funds and external donations. Exchanges and payment providers are expected to replay or reimburse affected deposits and withdrawals. Zano said limited liquidity may have prevented the attacker from selling the counterfeit tokens at scale. The Zano exploit is bearish for ZANO in the short term because it undermines confidence in token supply integrity, gateway security and blockchain immutability. Traders should monitor exchange reopening notices, liquidity, wallet balances, compensation progress and the project’s root-cause analysis. Longer term, confidence may depend on whether Zano delivers a credible security fix and completes the rollback recovery.

Crypto hacks caused $766.49 million in losses in September, according to PeckShield, up 462% from August’s $136.3 million and making it the worst month for crypto theft in 2026. Two incidents accounted for most of the damage: the Bitget breach, estimated at about $387 million, and the Liquid Network theft of roughly $320 million. Bitget said attackers compromised a backend wallet-infrastructure system and manipulated transaction data without obtaining private keys or accessing cold wallets. The exchange plans to reimburse users through a protection fund worth more than $464 million. BGB traded near $1.98 on 1 October, down 2.62% over seven days. The Liquid Network incident involved about 4,000 BTC withdrawn from a federation wallet. The attacker later said funds would be returned after network nodes were patched. Around $285 million was reportedly recovered, although earlier estimates put the unrecovered amount near $47 million. SlowMist said North Korea-linked hackers were laundering stolen Bitget funds through Chainflip, converting the proceeds to BTC and using CoinJoin to obscure transaction paths. Chainflip has attempted to block related deposits. Excluding Bitget and Liquid, 53 other crypto hacks caused about $59 million in losses, below August’s level. Total crypto theft in 2026 reached roughly $2.68 billion across 656 incidents, highlighting continuing exchange, wallet, bridge and smart-contract security risks. The incidents may weigh on trader confidence in the short term and increase demand for stronger custody and compliance controls.

Crypto job postings surged to 1,241 in September, up from 886 in August and 382 in July, according to CryptoJobsList data cited by CoinDesk. The total was more than twice January’s 573 listings and came as the number of hiring companies rose to 125, compared with 77 in August and 107 in July. Crypto job postings were concentrated in finance, engineering and trading. Stablecoins, artificial intelligence, security and compliance were also major hiring areas. Bitcoin, Ethereum and Solana were the most requested blockchain skills. Applications moved in the opposite direction, falling from 25,700 in July to 24,631 in August and fewer than 20,000 in September. The divergence may point to tighter talent supply, weaker candidate interest or more specialised requirements. The hiring recovery is a positive signal for crypto infrastructure, finance and compliance firms, but it is not a direct price catalyst for BTC, ETH or SOL. Traders are likely to view the development as neutral unless it is supported by stronger investment, revenue growth or token-specific catalysts.

BlackRock is expanding tokenization from individual securities to complete onchain portfolios through a partnership with Ondo Finance. It developed three tokenized strategies focused on high income, diversified growth and high growth. Each strategy is represented by a transferable token, giving investors exposure to multiple underlying assets through a single blockchain-based position. The BlackRock tokenization initiative could make portfolio allocations more portable and transparent. In the future, the tokens may move between wallets and platforms, serve as collateral for lending, or connect with decentralised finance products. Automated software could also adjust allocations as market conditions change. Broadridge estimates that traditional model portfolios held about $9.8 trillion in assets as of June. The development reflects a broader shift from tokenized individual securities to tokenized investment portfolios. Bitwise has launched a similar automated portfolio model with Coinbase and Glider, although its tokenized stocks remain in users’ wallets while software manages target allocations. For crypto traders, the launch may support long-term demand for ONDO and tokenized real-world asset infrastructure, but it is not an immediate price catalyst. Wider adoption will depend on liquidity, regulatory clarity, custody, prime-brokerage infrastructure and the number of assets brought onchain. Industry executives also see potential applications across private credit, private equity, crypto and international assets, particularly when combined with artificial intelligence and stablecoins.

Tavus says its Griffin AI model achieved a major video Turing test result. In a one-minute live call, 26 of 54 participants, or 48%, believed Griffin-Lite was human. Tavus’s previous system convinced only one of 41 participants, or 2.4%. The company’s study was not independently verified and does not follow a standard Turing test protocol. Griffin AI combines video, audio, speech, facial expressions and pauses for real-time interaction. On NVIDIA’s VideoFDB benchmark, Griffin-Lite scored 3.73 out of 5 for perception and 3.83 for response generation. Human reference scores were 4.20 and 3.92. Reported latency ranged from a 0.43-second average on H100 chips to benchmark median response times of 1.89 seconds for generation and 2.23 seconds for perception. The model could support customer service, coaching and digital assistants, but it also raises risks involving deepfakes, impersonation and video fraud. Griffin-Lite remains limited to trusted testers while Tavus develops disclosure and safety controls. For crypto traders, the development is relevant to identity verification, cybersecurity and video-based scams, but it provides no direct catalyst for cryptocurrency prices.

Chainalysis said its AI tools helped investigators trace the $387 million Bitget hack and cut cross-chain bridge reconciliation from more than 20 hours to under 10 minutes. Human analysts set the rules, reviewed the results and directed the investigation. The firm attributed the Bitget hack to North Korean actors. Within three hours of the 24 September breach, 23 transfers moved about $387 million from Bitget. The stolen assets were distributed across Ethereum (49.7%), XRP (40.8%), Zcash (7.6%) and Tron (1.8%). Investigators traced stolen XRP through a cross-chain liquidity protocol that paid out Bitcoin, linking the funds to attacker-controlled BTC addresses. Bitget said attackers compromised a backend system and forged withdrawal commands, while cold wallets and private keys remained secure. The exchange raised its loss estimate to $387.5 million and later restored Bitcoin, Ether and USDT withdrawals. Bitget is offering separate 5% rewards for qualifying assistance that freezes or recovers the assets. The Bitget hack highlights rising exchange security, bridge risk and crypto crime concerns. Chainalysis has labeled related addresses and is sharing intelligence with exchanges, issuers and law enforcement. The direct impact on broad crypto prices is likely limited, but affected assets and crypto-security tokens could face short-term volatility.

Genius has ended Season 2, distributing 173 million GP, below the planned 200 million. The allocation represents up to 6.055% of GENIUS’s total supply, while the unused 0.945% will be automatically burned. GENIUS users can request a full refund of platform fees from 00:00 on 7 October to 00:00 on 12 October. Users who claim refunds will forfeit their Season 2 GP allocation, and the corresponding GENIUS tokens will be repurchased and permanently burned. Users who do not request refunds will have a second five-day window from 12 October. They can unlock their GENIUS allocation immediately with an 85% deduction or lock it for 24 months to preserve 100% of the airdrop allocation. The refund, token burn and unlock choices could affect GENIUS supply, selling pressure and short-term volatility.

A violent UK crypto robbery in Solihull has highlighted the growing risk of wrench attacks against digital-asset holders. In December, three masked intruders allegedly broke into a couple’s home, beat the husband with hammers for about 45 minutes and threatened his seven-month-pregnant wife. They forced him to unlock his phone and, through a live video call, identified a wallet before directing a transfer worth hundreds of thousands of pounds. The attackers also stole several Rolex watches. Police are investigating, and Crimestoppers is offering a reward of up to £10,000 for information. The crypto robbery is unlikely to result in a recovery because blockchain transfers are difficult to reverse once funds reach an attacker-controlled wallet. The incident adds to wider concerns about physical coercion targeting self-custodied assets. Chainalysis recorded $30 million stolen in violent crypto robberies by June 2026, while CertiK identified 52 verified wrench attacks involving about $124.1 million in exposure during the first half of the year. The datasets measure different factors and should not be combined. For traders and long-term investors, the crypto robbery reinforces the value of limiting public disclosure of holdings, separating funds, using multisignature or hardware wallets and setting withdrawal controls. These measures can reduce operational risk, but no technical setup can fully eliminate the threat of physical coercion.

Bitcoin price volatility intensified after a softer-than-expected US jobs report initially pushed BTC from about $86,000 above $87,000. The rally was quickly rejected, sending Bitcoin below $84,000 and causing a decline of more than $3,000 within hours. BTC later stabilised near $84,500 after repeatedly defending support around $83,000, but it remains below the $85,000 resistance level. Nearly $600 million in crypto positions were liquidated over 24 hours, with long positions accounting for almost all recent hourly liquidations. The largest single liquidation was close to $12 million on Binance. Bitcoin’s market capitalisation stood at about $1.69 trillion, while its dominance rose to 59%. Ethereum fell below $2,700, XRP declined 3.5% to below $1.50, and Zcash dropped more than 5%. Dogecoin, Chainlink, Cardano, RAIN, Stellar and NEAR also weakened. Total crypto market capitalisation fell nearly 2%, or about $80 billion, to approximately $2.88 trillion. QNT and NIGHT bucked the broader crypto market correction, each gaining about 13%, with QNT moving above $260 and NIGHT reaching $0.50. Despite supportive signals from inflation and the US labour market, the Bitcoin price rejection and heavy liquidations show that traders remain cautious and that leveraged positions face continued downside risk.

The SEC has proposed a crypto custody framework for registered investment advisers, investment companies and business development companies. The proposal would allow conditional self-custody when no permitted custodian supports a specific digital asset. Advisers would have to document the decision and reassess it quarterly. The crypto custody framework would also allow state trust companies to hold digital assets, subject to verification of their legal authority and written safeguards against theft, loss, misuse and misappropriation. These safeguards would require annual reviews. The SEC said existing Investment Advisers Act and Investment Company Act rules were designed for traditional assets and have not kept pace with crypto markets. The proposal also includes updated recordkeeping requirements and is intended to reduce custody barriers to crypto investment advice and fund strategies. It is not final. A 60-day public comment period will begin after publication in the Federal Register. The plan follows the SEC’s broader Regulation Crypto Assets proposal, while wider legislation known as the Clarity Act remains stalled. For crypto traders, the immediate price impact is likely limited because implementation remains uncertain. Over the longer term, clearer crypto custody rules could support institutional participation and on-chain markets in the United States. Traders should monitor Federal Register publication, SEC proposals and industry responses.

NEAR Intents recovered the full $3.8 million stolen in a 1 October exploit, less than 16 hours after general manager Alex Shevchenko gave the attacker 48 hours to return the funds. The recovery ended the investigation, although the team has not released a detailed incident report. Blockchain data shows about 34.59 BTC, worth roughly $2.95 million at the time, was sent to an official Bitcoin address. Only small amounts reached the disclosed BNB and Ethereum addresses, while the Solana address received almost nothing. Shevchenko said the remaining funds may have been returned through another channel. The exploit reportedly involved a vulnerability in Omni’s deposit and withdrawal infrastructure and smart contracts. Losses were limited to USDT on BNB Chain. NEAR’s mainnet and NEAR token were not affected, and the vulnerability was fixed within an hour of discovery. NEAR Intents said its SHIELD AI monitoring layer and internal investigation helped identify the attacker. The platform is urging researchers to use its bug bounty programme, which offers up to $300,000 for critical smart-contract flaws and $100,000 for cross-chain bridge vulnerabilities.