Menu
MCAP $2.83T ▼-2.68%
BTC $83,768 ▲+0.93%
ETH $2,550 ▲+1.62%
BNB $753.21 ▲+0.69%
XRP $1.410 ▲+0.38%
SOL $111.69 ▲+1.52%
DOGE $0.0868 ▲+0.97%
ADA $0.250 ▲+0.15%
TRX $0.3306 ▲+0.09%
LINK $13.35 ▲+2.64%
AVAX $10.91 ▲+5.56%
HYPE $86.62 ▲+1.06%
DOT $1.240 ▼-1.05%

Cryptocurrency Security

Every Crypto Coin Show story tagged Cryptocurrency Security, newest first, alongside the guides and interviews that give it context.

Guide: Perpetual futures, funding rates and liquidations,Guide: US crypto tax rules explained (2026):Newsletter
From the interview archive
Lightning Development Kit patches theft flaw affecting unpatched Bitcoin wallets
Bitcoin

Lightning Development Kit patches theft flaw affecting unpatched Bitcoin wallets

, , , , , , , ,

BitcoinCrypto Coin Show News Team·October 11, 2026·3 min read

Lightning Development Kit maintainers shipped security releases that close a flaw letting a dishonest channel partner steal a forwarded Bitcoin payment by lying after a reconnection. Any wallet or payment app built on an unpatched copy of the toolkit remains exposed until its developers rebuild and redeploy it, which matters because LDK underpins custodial and non-custodial Lightning products that institutions use for instant Bitcoin settlement.

  • LDK’s v0.2.7 and v0.1.13 security releases, dated Oct. 1, fix a reconnect flaw across both the 0.2 and 0.1 code branches.
  • Before the fix, a peer could falsely claim it never received an acknowledged update, tricking LDK into signing an untracked commitment transaction.
  • Version 0.2.7 separately fixes an LSPS2 just-in-time channel bug that could have forced a liquidity service to cover an attacker’s shortfall from its own funds.
  • Oct. 1 date both patched branches, v0.2.7 and v0.1.13, were released
  • Sept. 13 prior LDK v0.2.6 disclosure

Lightning Development Kit maintainers released v0.2.7 and v0.1.13 on Oct. 1, 2026, patching a flaw that let a malicious channel partner steal the value of a forwarded Bitcoin payment by lying after reconnecting, according to the Oct. 9 newsletter from Bitcoin Optech. LDK is a software toolkit, not a standalone program, that developers compile directly into wallets and payment apps to run Bitcoin’s Lightning Network, the layer-two protocol used for fast, low-fee transactions off Bitcoin’s base chain.

Because LDK ships as embedded code rather than a shared daemon, the fix only protects applications whose developers actually rebuild and redeploy it. CryptoSlate detailed that distribution problem in its report on the vulnerability.

A Reconnect Lie Let Peers Sign a Commitment Transaction LDK Never Tracked

The attack begins with a channel peer acknowledging a payment update, then reconnecting and falsely claiming it never received that acknowledgment. Before the fix, LDK could respond by signing a conflicting commitment transaction, the record of a channel’s agreed state that can be broadcast to settle it on Bitcoin’s blockchain. The flaw, tracked in PR 5057, meant that newly signed transaction never got recorded by LDK’s channel monitor, the component responsible for tracking a channel’s on-chain claims.

That gap turned a forwarded payment into a potential loss for the operator caught in the middle. A malicious sender could confirm the false transaction on-chain, let the payment settle with the downstream recipient, and then reclaim the incoming payment contract once it expired, even though the forwarding node knew the secret needed to claim it.

The forwarding node would have paid out downstream without recovering the matching incoming funds. The fix now permits retransmission only while an acknowledgment remains genuinely outstanding and force-closes the channel if a peer claims a confirmed update was missed.

Version 0.2.7 Also Closes an Overpayment Gap for Liquidity Providers

Alongside the reconnect fix, v0.2.7 addresses a separate issue in LSPS2, a specification that lets a liquidity service provider open a channel on the fly when it receives a payment larger than a user’s existing capacity, a mechanism commonly used for instant onboarding in mobile and custodial Lightning wallets. An intercepted payment could misrepresent its own amount, pushing the service to open a channel and forward more Bitcoin than it actually received. The provider would then have to cover the shortfall from its own balance, and PR 5042 adds the missing amount check.

The v0.1.13 release notes list only the shared reconnect fix, not the LSPS2 fix, leaving apps on that branch exposed to the liquidity-service issue if they use the feature. The PR 5042 notes also flag that payment contracts queued under a prior version still carry unvalidated amounts, so teams must account for those pending contracts separately from simply upgrading the library.

LDK’s Embedded Design Means Every App Has to Rebuild Itself

LDK’s own architecture documentation explains why patching is uneven across the ecosystem: the SDK is compiled and executed inside each application rather than run as a shared daemon, leaving developers to choose their own storage, wallet and blockchain-monitoring code. That environment-agnostic design is what let CryptoSlate’s prior v0.2.6 report on Sept. 13, 2026 describe a separate set of splice-fee and state-loading bugs just weeks before this disclosure.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Ledger confirms hidden implant in CryptoBilis victim's wallet as losses near $93 million
Blockchain

Ledger confirms hidden implant in CryptoBilis victim’s wallet as losses near $93 million

, , , , , , , , , ,

Legal & CrimeAshton Addison·October 10, 2026·6 min read

Ledger confirmed on Saturday that at least one wallet belonging to a victim of the CryptoBilis drain had an unauthorized hardware implant inside it. The finding backs up the photos former Mt. Gox CEO Mark Karpelès posted a day earlier, but Ledger still has not said the implant caused the losses, which on-chain trackers now put between $72 million and $93 million.

  • Ledger Support said on October 10 that one impacted user’s device contained an unauthorized hardware implant, the first company confirmation of physical tampering in the case.
  • CryptoBilis, an authorized Ledger reseller for Malaysia, Indonesia and the Philippines, has stopped selling all hardware wallet inventory, not only Ledger units, until the investigation ends.
  • Karpelès’s teardown shows a cellular implant hidden behind the screen that reads the recovery phrase off the display during setup, a design that can still pass Ledger’s Genuine Check.
  • $72M to $93.4M Range of suspected losses across independent on-chain estimates; Ledger has confirmed none of them
  • 471 addresses Suspected victim addresses counted by tracker Yfarmx in its $93.4 million estimate
  • 90 days Purchase window Ledger flagged for CryptoBilis buyers

Ledger confirms an implant, not yet a cause

Ledger’s situation update on Saturday moved the story from rumor to documented tampering. The company said one of the impacted users’ devices held an “unauthorized hardware implant,” that it is contacting affected customers, and that anyone with information should reach it through its bounty program. It thanked the volunteer security response group SEAL 911 for help and said it is working with authorities.

Ledger also said it has no sign that its own security infrastructure, systems or services were compromised, and that it is building stronger anti-tampering measures. What it has not said is just as important: how the implant got into the device, how many units carry one, and whether the implant explains every drained wallet. Bitcoin.com News and others noted that Ledger has not verified any of the loss totals.

The update came one day after Ledger first asked CryptoBilis to pause sales and shipments on October 9. CCS covered that first pause here. CryptoBilis has since widened the halt to its entire hardware wallet inventory, which matters because the reseller also lists other wallet brands.

What Mark Karpelès found inside the Nano X

Karpelès, who ran the Mt. Gox exchange before its 2014 collapse, posted photos on October 9 of a Ledger Nano X circuit board next to the implant he said was hidden inside it. He asked anyone affected to open their device and send him pictures so he could check for the same hardware.

According to Karpelès, the implant packs a microcontroller, an LTE cellular modem, an antenna and an eSIM into the foam pad that normally sits behind the screen. The microcontroller is wired to the SPI bus, the internal line the Nano X uses to send text to its display. That placement is the whole trick. A hardware wallet never lets the seed phrase leave its secure chip, except in one place: the screen, during first setup, when the owner writes the 24 words down. An implant that copies what the screen shows can capture the phrase at that moment and send it out over a mobile network, with no internet connection to the computer needed.

Because the genuine Ledger secure element is left untouched and still generates a real seed, the device can pass Ledger’s Genuine Check. Ledger’s own threat model documentation says that check verifies the secure element and cannot guarantee that the rest of the hardware is unmodified. Physical signs that a unit was opened include a stray antenna wire, a battery that looks smaller than normal, or a missing pad behind the screen.

Karpelès has since published a research page collecting his findings and said his next steps are to analyze each SIM card and extract the firmware from each microcontroller. He also asked CryptoBilis to open unsold units from its stock to see whether they carry the same part. On October 9 it was not yet established that the device he photographed came from CryptoBilis; Ledger’s Saturday confirmation is the first link between an implant and a CryptoBilis victim.

How big the losses are, and where the money went

The totals keep moving because each tracker counts a different set of addresses, and some may overlap:

  • Analyst Specter put losses above $86 million across Bitcoin, Ethereum and TRON.
  • Investigator tanuki42 traced more than $72 million to suspected theft addresses.
  • Arkham put the figure above $80 million; SlowMist’s MistTrack said it was approaching $90 million.
  • Yfarmx counted $93.4 million across 471 addresses, and Bitquery about $92.9 million across 311 unique addresses.

MistTrack reported that Tether froze a significant amount of USDT at addresses tied to the incident, which may limit how much of the stablecoin portion the attackers can move. Bitcoin and ETH have no equivalent freeze.

One case shows how the attack plays out. Lookonchain flagged a trader who bought a new Ledger from CryptoBilis and moved 80 BTC onto it on September 29. Public mempool records show all 80 coins, roughly $6.6 million, leaving in a single transfer at 05:54 UTC on October 9, ten days later. Another user, Edward Winz, said publicly that he lost about $1 million.

A known attack, now through an official channel

Hardware implants in Ledger devices are not new. Security researcher Joe Grand reverse engineered a Nano X implant found in a secondhand unit bought in Thailand, where the battery had been shrunk to make room for a small board that tapped the same display interface. Earlier cases involved USB drives soldered inside devices and fake units with swapped chips sold on marketplaces.

The difference this time is the sales channel. CryptoBilis was listed in Ledger’s official reseller directory, the place buyers are told to go precisely to avoid tampered stock. Binance co-founder Changpeng Zhao said early on that the evidence pointed to a supply chain attack localized to one vendor. Posts circulating on X allege that CryptoBilis changed ownership earlier this year; CCS has not been able to verify that claim, and neither Ledger nor CryptoBilis has addressed it.

Many replies to Ledger’s update asked whether victims will be compensated, arguing that an authorized reseller is Ledger’s responsibility. Ledger has not answered that question.

What CryptoBilis buyers should do now

  • Not set up yet? Do not initialize the device. Ledger’s guidance covers any unit bought from CryptoBilis in the last 90 days.
  • Already set up? Ledger advises moving funds to a new Ledger signer bought directly from the manufacturer, with a brand new seed. Do not reuse the old 24 words on the new device, because the old phrase may already be in someone else’s hands.
  • Bought from another reseller? There is no sign so far that other channels are affected, but checking for signs the case was opened costs nothing.
  • Watch for scams. Ledger will never ask for your recovery phrase. Expect fake “Ledger support” messages and refund offers to target victims.

The CCS read. This is a supply chain failure, not a broken Ledger chip, but that distinction offers little comfort to people who did everything right and bought from an authorized seller. The implant works because it attacks the one moment self-custody depends on a human reading a screen. Buying direct from the manufacturer just went from good advice to the only advice, and hardware makers will now be judged on how well they police their own reseller lists and whether they stand behind buyers when those lists fail.

Next to watch: whether Ledger confirms more implanted devices or ties the implant to the full loss figure, results of Karpelès’s SIM and firmware analysis, whether CryptoBilis explains how tampered units entered its stock, any recovery or freeze of the Bitcoin and ETH, and whether Ledger offers affected buyers compensation or replacement devices.

Get this in your inboxThe Crypto Coin Show newsletter covers the security, policy and market moves crypto investors need to know.
Subscribe
Ledger halts Malaysian reseller after $86 million in suspected wallet losses
Blockchain

Ledger halts Malaysian reseller after $86 million in suspected wallet losses

, , , , , , ,

BlockchainCrypto Coin Show News Team·October 9, 2026·3 min read

Ledger halted Southeast Asian reseller sales after reports that buyers lost $86 million from hardware wallets, but the company has not confirmed the figure or explained whether devices were tampered with. For institutional custody and portfolio managers, this incident underscores supply chain risks beyond official channels and parallels a 2024 case involving counterfeit units engineered to steal seed phrases at scale.

  • Ledger asked CryptoBilis, a Malaysian reseller operating in Southeast Asia, to pause all sales and shipments after loss reports surfaced Friday (October 9).
  • Analyst Specter traced suspected losses to hundreds of victim wallets across Ethereum, TRON and Bitcoin, with about 211 BTC in tracked addresses remaining stationary as of 13:44 UTC Friday.
  • The incident echoes a 2024 supply chain attack involving fake Ledger devices with replaced security chips that transmitted PINs and seed phrases to attacker servers.
  • $86M Estimated losses from CryptoBilis device sales, per analyst Specter
  • 211 BTC Combined holdings in three Bitcoin addresses Specter identified
  • 90 days Timeframe for affected purchases Ledger warned to move funds

Ledger told buyers who purchased from CryptoBilis within the past 90 days not to initialize their devices, and instructed those who had already done so to transfer funds to a new device with a fresh seed phrase. The pause, announced by Ledger Support on Friday, covered only this single Southeast Asian reseller. Ledger has not disclosed whether the devices themselves were altered, how funds left wallets, or whether the $86 million figure reflects confirmed losses or suspected activity on blockchain ledgers.

Counterfeit devices with swapped security chips preceded this incident

In April, a cybersecurity researcher documented counterfeit Ledger hardware wallets sold through a Chinese marketplace that sent PINs and seed phrases to attackers. The fake devices spoofed Ledger’s firmware to version 2.1, a release that does not exist in Ledger’s official lineup, and transmitted every PIN and seed phrase entered on the device to attacker-controlled servers in plaintext.

Ledger’s Genuine Check passed legitimate devices but CryptoBilis purchases remain unverified

Ledger’s built-in cryptographic Genuine Check successfully identifies counterfeit hardware when run through the legitimate Ledger Live application downloaded from ledger.com. However, the 2024 counterfeit operation used a cloned phishing website and trojanized app that always returned a false positive result, ensuring victims never received a warning signal.

The current CryptoBilis situation leaves open whether devices were altered before shipment, whether buyers received legitimate hardware sold through an unauthorized channel, or whether the drain involved a separate attack vector such as supply chain interception or compromised firmware during distribution.

Ledger said it will share updates as its investigation progresses, but has not specified a timeline for conclusions or whether it will release a forensic analysis of affected units.

The CCS read. Institutional holders should assume that any hardware wallet purchased outside official channels or verified resellers carries elevated risk of tampering or firmware compromise. The CryptoBilis case expands the attack surface beyond direct counterfeit manufacturing to include legitimate resellers whose fulfillment or warehousing may have been breached. Custody providers and portfolio managers relying on self-hosted cold storage should audit procurement policies and confirm device provenance.

Ledger has not announced a public audit of CryptoBilis’s supply chain, a recall mechanism for affected buyers, or an API-level defense to flag devices with suspicious transaction patterns. Institutional users should watch for Ledger’s investigation conclusion, for which no timeline has been specified, and any guidance on whether the losses resulted from hardware tampering, firmware compromise, or operational failure by the reseller itself.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Europol flags $502 billion in Bitcoin vulnerable to quantum computer theft
Bitcoin

Europol flags $502 billion in Bitcoin vulnerable to quantum computer theft

, , , , , , , ,

BitcoinCrypto Coin Show News Team·October 7, 2026·4 min read

Europol said Wednesday that Bitcoin wallets with exposed public keys have no cryptographic fix once quantum computers become powerful enough to break them, a narrow but concrete warning aimed at individual holders rather than the blockchain itself. The agency’s cybercrime unit puts a figure on that exposure: roughly $502 billion in BTC already shows the public key an attacker would need, held in reused addresses, exchange custody and early wallet formats.

  • Glassnode counted 6.04 million BTC exposed as of May, equal to 30.2% of all bitcoin ever issued.
  • Quantum-resistant signatures run 10 to 120 times larger than today’s, which Europol warns could clog blocks and push up fees.
  • A full migration to quantum-safe wallets would need at least 76 days of Bitcoin’s entire network capacity, one study Europol cites found.
  • 30.2% share of all bitcoin supply Glassnode flags as quantum-exposed
  • $502B value of exposed BTC at bitcoin’s roughly $83,050 price
  • 300 days network capacity needed to spread a full wallet migration

Every Bitcoin wallet runs on two linked codes: a public key that anyone can see on the blockchain, and a private key that authorizes spending. Europol’s cybercrime unit said a sufficiently powerful quantum computer could work backward from an exposed public key to derive the private one, letting an attacker drain the wallet, according to a report first detailed by BeInCrypto. The hashing that chains Bitcoin’s blocks together is far harder to break, Europol’s report found, so the threat sits with individual holders rather than the network’s consensus layer.

Glassnode Counts 6.04 Million Exposed Bitcoin, 30.2% of Supply

Analytics firm Glassnode found that 6.04 million BTC carried an exposed public key as of May, equal to 30.2% of all coins issued. At bitcoin’s price near $83,050, that stash is worth approximately $502 billion, a sum comparable to the market capitalization of a major global bank.

Most standard Bitcoin addresses show only a scrambled fingerprint of the public key, not the key itself, so those coins are not counted as exposed. That protection disappears the moment an address sends a transaction, which reveals the public key and strips cover from any coins left behind at that address.

Exchange-held coins land in the exposed group more often than self-custodied ones, Glassnode found, because platforms tend to reuse deposit addresses across many transactions.

Taproot, the address format Bitcoin added in 2021, and Satoshi Nakamoto’s earliest coins both show the public key by default rather than a hashed version of it, putting both the newest and oldest corners of the network at risk.

Europol Rejects Claims That Quantum Computing Breaks Crypto Outright

Europol’s warning is deliberately narrow. The agency rejects broader claims that quantum computing will collapse cryptocurrency, framing the risk instead as falling on holders who have exposed their keys, not on the protocol as a whole.

Once a public key is exposed, it remains permanently vulnerable to future quantum decryption.

Europol, cybercrime unit report

A second Europol report describes a tactic called harvest-now-decrypt-later, where an attacker copies encrypted data or exposed public keys today and waits for quantum machines capable of cracking them later. Europol said it found no clear evidence this is happening at scale against crypto holdings, but the strategy means exposure recorded today could still be exploited years from now even if no quantum attack exists yet.

Wallet Migration Would Need 76 to 300 Days of Network Capacity

Europol’s cited research estimates that migrating all exposed coins to quantum-safe wallets would consume at least 76 days of Bitcoin’s entire transaction capacity if done all at once. Spreading that migration across a quarter of every block instead stretches the timeline to roughly 300 days, a tradeoff between network congestion and migration speed.

Europol set no date for when a quantum computer capable of this attack might exist. It argues that upgrading a network with no central operator takes years, and noted that some Bitcoin developers already back proposals to freeze coins in vulnerable wallets ahead of any attack.

The larger signature sizes Europol flags, up to 120 times current size, raise the same engineering tension already visible in other parts of crypto infrastructure, where security upgrades compete with throughput. The Ethereum Foundation’s recent proposal for native transaction assertions addresses a related signing-failure risk, while Bitcoin’s own node software has moved incrementally on security, as in Core Lightning’s v26.06.9 release earlier this year.

The CCS read. Custodians and exchanges carry more of this exposure than retail holders who never reuse addresses, since pooled deposit wallets are precisely the reused, always-public-key-visible structures Glassnode flags. Institutional custody providers should treat key rotation and address hygiene as a near-term operational item, not a theoretical one tied to a quantum computer that does not yet exist.

Europol has not set a timeline for when migration standards or network-level freezes might be proposed to Bitcoin’s developer community, leaving custodians and exchanges to decide on their own whether to rotate exposed addresses before any quantum capability materializes.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Bitget’s September hack drained $388 million from hot wallets, exchange confirms
Exchange News

Bitget’s September hack drained $388 million from hot wallets, exchange confirms

, , , , , , ,

Exchange NewsCrypto Coin Show News Team·October 6, 2026·4 min read

Bitget confirmed in a security update published through its own Academy channel that the September 24 hack drained a final verified $388 million from its hot and warm wallet infrastructure, up from the roughly $387.5 million first estimated. The disclosure matters to institutional counterparties because it tests whether the exchange’s protection fund and outside capital can absorb a nine-figure loss without disrupting custody, withdrawals or trust in its reserve reporting.

  • Bitget’s final verified tally rose to $388 million, covering 12 wallet addresses across its hot and warm tiers.
  • The exchange replenished its protection fund above $300 million and completed phased withdrawal resumption on October 2.
  • CEO Gracy Chen told BeInCrypto the company holds more than $1 billion in capital outside that fund, but could not explain how attackers learned Bitget’s internal systems.
  • $388M final verified theft, versus Bitget’s initial $387.5 million estimate
  • 131% reserve ratio across 19 assets in Bitget’s 47th proof of reserves
  • $1.07M frozen by trackers, just 0.28% of the stolen total

Bitget’s security update, confirmed as of October 6, puts the final figure at $388 million, slightly above the roughly $387.5 million first reported in the days after the attack.

The exchange said the intrusion began at 18:31 UTC on September 24, when an attacker exploited a vulnerability in a third-party security product to obtain high-level internal credentials, then used them to send fraudulent withdrawal commands that bypassed existing risk controls.

Chainalysis has attributed the theft to North Korean actors, according to its analysis, consistent with earlier state-linked exchange breaches. Crypto Coin Show first covered the breach when Bitget paused withdrawals and pointed to its replenished protection fund above $300 million days after the attack.

Bitget’s own account, detailed in its Academy update, rules out a private-key compromise and says cold wallets across all chains were untouched. The incident was limited to a portion of its hot and warm wallet infrastructure, the exchange said, in a case first detailed by BeInCrypto’s interview with CEO Gracy Chen.

Bitget Rules Out Private-Key Breach, Confirms 12 Wallets Hit

Bitget’s timeline shows its reconciliation system detected a discrepancy shortly after the first unauthorized transfers, triggering an automatic risk-control block on withdrawals across the platform.

Because private-key compromise had not yet been ruled out at that point, the wallet team moved remaining funds to cold storage and shut down signing services entirely, isolating withdrawal-related access while the security team traced the attack path.

Twelve wallet addresses tied to hot or warm infrastructure were affected, spanning assets including XRP, ETH, USDT, ZEC, ATOM, USDC, USD0, XAUt, BNB, AVAX, TRX, ALGO and TIA. Bitget Wallet, a separate non-custodial product that keeps user assets onchain under their own control, was not touched.

Mandiant and SlowMist Confirm Vendor Compromise, Withdrawals Fully Restored October 2

Independent investigation reports from Mandiant, part of Google Cloud, and SlowMist became available on September 30, and both broadly aligned with Bitget’s internal findings. The firms identified the compromise of a third-party security product as the mechanism that ultimately enabled unauthorized access to the exchange’s wallet environment, according to the bitget.com security update.

Resumption followed in stages. ETH withdrawals returned first across Ethereum, BSC, Arbitrum, Base and Optimism, followed by USDT across Ethereum, BSC, Solana and Tron once the protection fund had been rebuilt to more than $300 million.

Bitget’s 47th proof of reserves update, released during the same window, reported a 131% overall reserve ratio across 19 covered assets, before the exchange completed the plan on October 2 by restoring remaining tokens, fiat and C2C services.

Internal login credentials were revoked and reissued, and critical operations now require multiple approvals. The affected third-party vendor was notified and the exploited functionality disabled, Bitget said, while asset tracing with industry partners continues.

Chen Says Bitget Can Absorb Another Loss, But Can’t Explain How Attackers Got In

In the interview, Chen said Bitget had used bitcoin from its protection fund to meet withdrawals before rebuilding it past $300 million, a level she called sufficient even though the September theft exceeded that threshold.

She added the company holds more than $1 billion in capital outside the fund and could absorb another loss of several hundred million dollars, though the interview did not establish how much of that capital is immediately liquid.

Asked how attackers acquired enough knowledge of Bitget’s systems to build a dedicated withdrawal tool, Chen said a preliminary internal investigation found no insider involvement but she had no explanation for the attackers’ access. She said the probe could take longer than Bybit’s prior wallet breach because Bitget’s systems were more complicated.

No system, no security infrastructure is perfect or impenetrable.

Gracy Chen, CEO of Bitget

Trackers show only about $1.07 million frozen so far, roughly 0.28% of the stolen total, with the amount actually returned undisclosed. Certora’s post-interview analysis estimated that roughly $238 million left Bitget’s wallets even after the initial withdrawal block, a gap that will test whether the new multi-approval controls can stop forged transfers once an alarm sounds.

Chen also rejected the idea that protocols such as THORChain bear chief responsibility for laundering stolen exchange funds, saying permissionless infrastructure has varying technical capacity to intervene, as other chains have shown with emergency security measures of their own. She said Bitget has held preliminary talks about an exchange-and-protocol alliance against such attacks, but acknowledged aligning competitors’ interests and disclosure limits remains unresolved, a tension not unlike the aftermath of the Humanity Protocol hack earlier this year.

The CCS read. We read this less as a security story than a counterparty-risk story. Institutional desks routing liquidity through Bitget care more about the untouched capital buffer than the vulnerability patch, because self-reported reserve ratios and fund replenishments reassure depositors without proving the new approval layers would catch a repeat forged-transfer bypass. Until an independent audit covers access controls, not just custody balances, confidence rests on Chen’s word alone.

Bitget says asset tracing and recovery remain ongo

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Bitcoin wallet Electrum patches Lightning backup flaw affecting fund recovery
Bitcoin

Bitcoin wallet Electrum patches Lightning backup flaw affecting fund recovery

, , , , , , , ,

BitcoinCrypto Coin Show News Team·October 2, 2026·4 min read

Electrum developers closed a Lightning Network backup flaw that left some Bitcoin wallet users unable to reclaim funds after a channel closed remotely, but the fix does not retroactively repair backups already saved to disk. For institutional holders running self-custodied Lightning infrastructure, the episode is a reminder that off-chain key management carries recovery risks that standard custody audits rarely test.

  • Electrum 4.8.2, dated Sept. 11, remains the newest build listed on Electrum’s official site as of Oct. 1.
  • Wallets using BIP39 seeds or imported extended private keys always generate non-deterministic Lightning keys and stay at risk.
  • Electrum-seed wallets are only safe if the wallet file was created in version 4.1 or later, not merely upgraded to it.
  • 4.8.2 patched Electrum release, still current three weeks after shipping
  • Sept.11 release date of the fix, versus Oct.1 confirmation it is current
  • v4.1 wallet-file version threshold separating safe from vulnerable seeds

Electrum’s development team merged a fix for a backup defect in its Lightning Network integration, according to the pull request on the wallet’s GitHub repository. Lightning is Bitcoin’s off-chain payment layer that lets users transact without posting every payment to the base blockchain, and anchor channels are a Lightning channel type designed to let either party adjust fees when closing a channel. The defect meant that if a peer force-closed an anchor channel remotely, certain saved backups lacked the private-key material needed to sweep the user’s share of the balance back onto the Bitcoin chain.

The fix shipped in Electrum 4.8.2, released Sept. 11, still the newest build on Electrum’s official website as of Thursday, Oct. 1. It does not rewrite old backups automatically.

The gap was first detailed by CryptoSlate, which traced the fix to two linked pull requests and Electrum’s own warning language for affected wallets.

PR 10851 Adds Missing Payment-Key Data to Channel Backups

The pull request extends Electrum’s channel backup format to include the payment_basepoint, the key material needed to claim an anchor channel’s to_remote output after the peer force-closes it. Without that field, an exported backup could record that funds existed in a channel but lacked the specific key required to sweep them.

The fix also hides the option to request a remote force-close whenever the backup on hand cannot sweep the resulting output, preventing a user from triggering a close their own backup cannot settle.

A companion change, PR 10851, repairs full wallet-file exports. Maintainer SomberNight noted in the pull request discussion that earlier exports deleted a randomly generated Lightning private key, so re-enabling Lightning from that backup would generate different keys entirely, leaving the old channel records useless for spending anchor-channel funds.

BIP39 and Imported-Key Wallets Carry the Permanent Risk

Electrum’s release notes set two conditions that must both apply for a backup to be affected: the wallet uses non-deterministic Lightning keys, and the backup involves an anchor channel. Wallets built from BIP39 seeds or imported extended private keys always fall into the non-deterministic category, because those keys cannot be regenerated from the seed the way Electrum’s native format allows.

Electrum-seed wallets are the exception, but only partially. Their Lightning keys are deterministic solely if the wallet file was created in version 4.1 or later; a file originally generated under 4.0.x stays non-deterministic no matter how many times the software itself is upgraded.

On desktop, Electrum’s wallet information panel now flags channels as non-recoverable from the seed, while the Android app warns in its channel-opening dialog before a user even funds a channel.

The distinction did not exist in earlier releases, where the seed was broadly assumed to regenerate all wallet keys, including Lightning ones. It leaves one question Electrum’s documentation does not resolve: users who have already discarded the original wallet file that produced an at-risk backup have no stated path back to a safe export.

Startup Warnings Push Users Toward Manual Re-Export

Electrum 4.8.2 fixes the backup logic for future exports, but it cannot repair files generated before the update. The documented remedy is a fresh export from the running wallet, and Electrum now displays a startup warning to any wallet it detects as affected under the two conditions above.

That remedy depends on still holding the original wallet data needed to generate a new export. Installing 4.8.2 changes the software, but it does not recover key material already lost along with deleted wallet files.

The CCS read. The bug matters less for its scale than for what it says about Lightning tooling broadly: recovery assumptions baked into seed phrases do not automatically extend to off-chain layers, a gap custodians building institutional Lightning rails, and projects pursuing related upgrades like post-quantum Lightning changes, will need to account for in their own backup audits.

Electrum has not said whether a future release will attempt to identify which specific users still hold funds in anchor channels opened before the Sept. 11 patch, leaving the re-export step entirely manual and voluntary, much as node operators faced after unrelated client updates like Geth’s v1.17.7 release required opt-in configuration changes rather than automatic fixes.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Crypto Coin Show
Bitcoin

MEXC doubles Guardian Fund Bitcoin holdings to 2,000 BTC

, , , , , , ,

BitcoinCrypto Coin Show News Team·October 1, 2026·3 min read

MEXC has added another 1,000 BTC to its Guardian Fund, the exchange’s publicly disclosed user-protection reserve, bringing total holdings to 2,000 BTC alongside 100 million USDT. The move matters to institutional investors because it signals exchanges are now competing on balance-sheet depth and on-chain verifiability of reserves, not just on fee structures or listings.

  • MEXC added 1,000 BTC to its Guardian Fund on Thursday, October 1, 2026, its second Bitcoin allocation to the fund this year.
  • The fund now holds 2,000 BTC plus 100 million USDT, up from an all-stablecoin launch of $100 million in June 2025.
  • MEXC is targeting $500 million in total fund size within two years of a May 2026 commitment to expand the reserve.
  • 2,000 BTC Guardian Fund’s current Bitcoin holdings, double May’s 1,000 BTC total
  • $500M two-year fund target, five times its $100M June 2025 launch size
  • 100M USDT stablecoin reserve held alongside the fund’s Bitcoin

MEXC disclosed the new allocation through publicly disclosed wallet addresses that allow independent verification of the fund’s holdings. The second Bitcoin allocation sits at a distinct address, separate from the wallet that received the first 1,000 BTC in May. The USDT portion of the reserve remains trackable on Etherscan.

MEXC Doubles Bitcoin Reserves to 2,000 BTC

MEXC’s Guardian Fund began as a $100 million reserve held entirely in USDT. In May 2026 the exchange committed to grow the fund fivefold, to $500 million over two years, and added its first 1,000 BTC alongside the existing stablecoin base.

Thursday’s addition doubles that Bitcoin position to 2,000 BTC. MEXC’s chief executive framed the move as part of an ongoing process rather than a one-time milestone.

Security and user protection is not a one-time investment. It is a continuous commitment. The risks facing digital asset platforms continue to evolve, and the resources behind user protection need to evolve with them. Adding another 1,000 BTC to the Guardian Fund reflects our long-term approach to building stronger protection for users and ensuring that the infrastructure behind that protection continues to grow with MEXC.

Vugar Usi, CEO of MEXC

Guardian Fund’s Model Breaks From Third-Party Insurance

The structural difference from conventional exchange insurance is the publicly traceable wallet model. Every MEXC-held BTC and USDT address tied to the fund is disclosed, letting users independently verify balances and transaction history on-chain rather than relying on an auditor’s attestation after the fact.

That transparency matters against a backdrop of recent exchange failures. MEXC’s publicly disclosed addresses are meant to let users check reserve adequacy before an incident occurs, rather than after.

What the disclosures do not specify is how compensation is triggered or capped per incident, or whether the Bitcoin component can be liquidated quickly enough to match the fund’s own “instant coverage” claim during a market stress event. MEXC has not published a claims policy alongside the wallet addresses.

MEXC Still Needs $400 Million More to Hit Its Target

At 2,000 BTC plus 100 million USDT, the Guardian Fund sits well below the $500 million target MEXC set for itself in May 2026, roughly 19 months from today’s deadline window. The exchange has not disclosed a pace for further allocations or whether future additions will again be Bitcoin-denominated.

The approach of building reserves through direct asset purchases, rather than pooled insurance, echoes a broader pattern of crypto-native firms holding Bitcoin on their own books, a strategy also visible in corporate treasuries such as Strategy’s bitcoin purchases. For MEXC, the open question is whether exchange-held reserves earmarked for user compensation should be treated differently from a corporate treasury position.

The CCS read. We read this less as a security headline and more as a balance-sheet signal. Exchanges that disclose wallet-level reserves are effectively competing with licensed custodians on verifiability, and MEXC’s bet is that users will choose venues where compensation capacity is checkable in real time rather than promised in a terms-of-service clause.

MEXC has not set a date for its next allocation toward the $500 million target, nor published the claims criteria that would govern how the Bitcoin portion gets deployed during an actual incident. Users and counterparties tracking the fund can watch the disclosed wallet addresses directly for the next move.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

On-chain trackers flag $176 million in outflows from Bitget wallets as users report withdrawal blocks
Exchange News

On-chain trackers flag $176 million in outflows from Bitget wallets as users report withdrawal blocks

, , , , , , ,

BlockchainCrypto Coin Show News Team·September 24, 2026·2 min read

Bitget is facing hack speculation after on-chain trackers flagged roughly $176 million in outflows from wallets tied to the exchange, with users simultaneously reporting blocked withdrawals. The exchange has issued no official confirmation or denial, leaving institutional counterparties and market makers to weigh exposure without a clear account of what happened.

  • Arkham data shows roughly $176 million moving from Bitget-labelled wallets within about a 20-minute window.
  • A single address, 0x770b…63Ee, repeatedly received ETH, AVAX, BNB, USDT, USDC and XAUT from the wallets.
  • Bitget has not confirmed a breach, even as users report withdrawal failures on the platform.
  • $176M value moved from Bitget-linked wallets
  • 20 min window in which multiple large transfers were executed in sequence
  • 6 distinct asset types swept into a single receiving address

Bitget users began reporting withdrawal problems Thursday (September 24, 2026), the same day on-chain monitors spotted the outflow. According to a report by BeInCrypto, the transfers originated from several Bitget-labelled hot wallets and at least one cold wallet, a combination that typically signals either an emergency sweep or a coordinated theft rather than routine settlement.

Arkham Flags $176 Million Leaving Bitget Wallets in 20 Minutes

Arkham Intelligence data cited in the report shows a series of large transfers executed in rapid succession, all within roughly 20 minutes. The funds came from multiple wallets the platform labels as its own, including at least one cold storage address.

BeInCrypto first flagged the activity in a post on X Thursday, noting more than $100 million had already left a Bitget cold wallet at the time of posting. That figure climbed as additional wallet movements were tracked, reaching the roughly $176 million total.

Address 0x770b…63Ee Collects Six Asset Types From Multiple Wallets

The receiving address, 0x770b…63Ee, appears repeatedly across the transfers, pulling in ETH, AVAX, BNB, USDT, USDC and XAUT from separate Bitget-linked wallets in a short span.

Several wallets sending different assets into one address within minutes is the pattern security researchers associate with exchange compromises, though exchanges do occasionally consolidate holdings during custody changes or wallet maintenance.

Some social media accounts claim a portion of the swept assets was converted into ETH, but that detail does not appear in the underlying data and remains unverified.

Bitget’s Silence Leaves Withdrawal Freeze Unexplained

Bitget has issued no statement confirming or denying a breach as of publication, even as user reports of blocked withdrawals continue to accumulate.

The gap between a 20-minute fund movement and an unexplained multi-hour withdrawal freeze is the detail institutional risk desks will watch most closely. Without proof-of-reserves confirmation or a wallet-by-wallet accounting from Bitget, counterparties have no independent way to size their exposure.

The CCS read. We read this as a custody-concentration problem more than a single exploit story. Any exchange whose hot and cold wallets can be drained through one receiving address within 20 minutes is signaling weaker segregation controls than institutional counterparties typically require, regardless of whether this specific incident turns out to be a hack or an internal transfer gone wrong.

Bitget’s next move is the open question: a confirmed breach disclosure, a proof-of-reserves statement, or continued silence while withdrawals remain frozen. Until the exchange addresses the $176 million in tracked transfers directly, institutional desks are likely to treat the platform as an elevated counterparty risk.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Attacker uses single wallet to steal $1.55M in FET and mint 408.5M NTX tokens
Legal & Crime

Attacker uses single wallet to steal $1.55M in FET and mint 408.5M NTX tokens

, , , , , ,

BlockchainCrypto Coin Show News Team·September 21, 2026·4 min read

A single wallet has been tied to both the $1.55 million theft of FET tokens from a SingularityNET-linked bridge and a subsequent mint of 408.5 million NTX tokens, according to an on-chain forensic report. The pairing indicates two separate compromised credentials were exploited in a coordinated operation rather than two unrelated incidents, a distinction that matters for any institution holding exposure to Fetch.ai, SingularityNET or NuNet infrastructure.

  • The attacker drained 8,721,530 FET, worth about $1.55 million at the time, from SingularityNET’s Ethereum-side bridge contract on Saturday (September 19).
  • Twenty-nine minutes later, a NuNet minter key dormant since March 2023 created 408,532,878 NTX, roughly 42% of NuNet’s documented token supply, and sent it to the same wallet.
  • Bitvavo suspended WMTX deposits and withdrawals on Sunday (September 20) and later halted trading, though the forensic evidence does not establish WMTX was compromised through the same mechanism.
  • $1.55M value of FET drained from the bridge contract in one transaction
  • 408.5M NTX minted, equal to about 42% of NuNet’s supply
  • 547.89 ETH attacker’s holdings by 1:10 UTC Sept. 20, worth roughly $1.44 million

The attack emptied TokenConversionManagerV3, the legitimate Ethereum-side lock-and-release component of SingularityNET’s bridge, according to an on-chain forensic report prepared by Athena and first detailed by CryptoSlate. Fetch.ai subsequently paused its AGIX-to-FET conversion service and Ethereum-side bridge contract as a precaution.

Fetch.ai said the affected infrastructure belonged to SingularityNET, primarily its Ethereum-Cardano bridge. The company’s own contracts and standard FET transfers kept operating throughout.

NuNet Minter Sends 408.5 Million NTX to the Same Wallet That Received Stolen FET

Investigators found that the FET loss traced to a compromised backend authorization key rather than a flaw allowing an attacker to bypass the bridge contract itself. The withdrawal transaction carried a valid signature from the address the contract was configured to trust, letting its conversionIn function release the full FET balance to an attacker-controlled wallet.

The NuNet activity supplied the strongest link tying the two thefts to one operation. At 20:50 UTC, 29 minutes after the FET withdrawal, the dormant minter key created 408,532,878 NTX and sent the entire batch to the wallet that had just received the stolen FET.

A later forensic pass pushed the timeline earlier still. At 19:36 UTC, 45 minutes before the FET drain, the NuNet minter sent 0.3667 ETH directly to the eventual receiving wallet, while a separate attacker-linked account moved 24.3 million NTX into it, and NTX sales through MetaMask’s swap infrastructure had already begun.

Uncapped Conversion Function Let Attacker Withdraw 8.72 Million FET in One Transaction

The contract’s own design amplified the damage once the key was compromised. Its 1 million FET transaction cap applied only to tokens moving out of Ethereum and was never enforced on conversionIn, letting the attacker withdraw all 8.72 million FET at once instead of in capped increments.

The signed authorization message also failed to bind a specific recipient, meaning a valid signature could redirect funds to any address the caller chose.

Liquidity constraints then shaped how much value the attacker could extract.

The stolen FET was swapped largely into Ethereum, while more than 217 million of the newly minted NTX was sold through decentralized venues before pools thinned; four later sales of 38.55 million NTX combined added only about 0.30 ETH, and a 10 million NTX transaction routed through Mayan Protocol produced roughly 940 USDT.

Bitvavo Halts WMTX Trading While Credentials Remain Unrotated

Bitvavo suspended WMTX deposits and withdrawals on Sunday (September 20) citing an active security incident, then temporarily halted trading, while telling customers their balances remained safe. Historical SingularityNET material shows WMTX shares infrastructure with FET and NTX through the Ethereum-Cardano bridge ecosystem, but the forensic report examined only the FET and NTX activity in detail.

Roughly five hours after the attack, the report’s first tracking window found the compromised FET bridge authorizer and NuNet minter credentials had still not been rotated or revoked. By then the FET bridge itself was empty and inactive.

That leaves a narrow but consequential window open. Refilling the FET conversion contract while the same authorizer stays trusted risks exposing fresh liquidity to another signed withdrawal, and NuNet faces a parallel risk as long as the compromised wallet retains minting authority.

The CCS read. The real exposure here sits with anyone treating AI-agent token bridges as fungible custody infrastructure rather than distinct trust boundaries. NTX holders absorbed dilution equal to nearly half the prior supply before any exchange froze trading, while WMTX got caught by association rather than evidence. Institutions building positions across the Fetch.ai, SingularityNET and NuNet complex should be pricing each bridge’s key-management practices separately, not as a single ASI Alliance risk.

Fetch.ai has not said when AGIX-to-FET conversions or its Ethereum-side bridge will resume, and Bitvavo has said WMTX trading and transfers stay restricted until it completes its own assessment. Whether either team has rotated the compromised authorizer and minter keys since the report’s five-hour snapshot remains the next verifiable marker for institutions tracking the affected infrastructure.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

ACINQ patches three Eclair flaws that let peers drain entire channel balances
Bitcoin

ACINQ patches three Eclair flaws that let peers drain entire channel balances

, , , , , ,

BitcoinCrypto Coin Show News Team·September 21, 2026·3 min read

ACINQ has patched three peer-triggered flaws in its Eclair Lightning Network software that could have let malicious counterparties siphon a node operator’s entire local channel balance straight into a Bitcoin transaction fee. The bugs, fixed in Eclair 0.14.3 on September 14, matter to institutional Lightning operators because they show fund-loss risk can hide in routine channel operations, not just in exotic attack scenarios.

  • ACINQ released Eclair 0.14.3 on September 14, 2026, patching three separate peer-triggered vulnerabilities.
  • The worst flaw let an adversarial peer propose a closing fee larger than the victim’s local balance, sending the entire balance to miners.
  • Eclair now caps automatically estimated channel-opening and splice fees at 50 satoshis per vByte by default.
  • 0.14.3 Eclair release version that patched all three flaws
  • 3 distinct peer-triggered vulnerabilities fixed in one update
  • 50 sat/vB new default ceiling on automatically estimated fees

ACINQ, the Bitcoin technology firm behind Eclair and the Phoenix Wallet, said the vulnerabilities could be exploited by malicious nodes during channel closures, splicing and on-the-fly funding, and it strongly recommended operators upgrade, according to reporting from CryptoSlate. The most severe of the three centered on cooperative channel closures, where Eclair was responsible for setting the closing fee.

Malicious Peer Could Route an Entire Balance to Miners

In the flawed version, an adversarial peer could propose a closing fee larger than the victim’s actual local channel balance. Eclair’s fallback fee-negotiation logic could accept that proposal, wiping out the operator’s output entirely and effectively converting the whole balance into a transaction fee collected by Bitcoin miners rather than returned to the node.

The patch closes that gap by rejecting any closing-fee proposal that exceeds an operator’s own configured maximum. Bitcoin Optech flagged 0.14.3 as a security release addressing closing, splicing and on-the-fly funding issues.

Splicing and On-the-Fly Funding Carried Separate Fund-Loss Paths

A second flaw could strand funds mid-splice, the process of resizing a channel’s underlying funding transaction without closing it. If Eclair signed its half of a splice first and the peer withheld its own signature, the node’s latest channel state could depend on a transaction the victim had no way to publish.

That same gap opened a route to losses on payments still in flight: an attacker could let the incoming leg of a relayed payment expire, then publish an older channel state and use the payment secret to collect the outgoing leg for itself. Eclair now force-closes using the newest state that is backed by a fully signed funding transaction, closing that timing window.

The third bug hit on-the-fly funding, a feature that opens a channel while forwarding a payment. A malicious wallet could time payment expiries so it collected the outgoing payment on-chain while the incoming leg expired, leaving the relay operator to eat the loss; Eclair now checks relay fees and expiry buffers before committing funds.

BTCPay Server Reports Separate Bots Probing Exposed LND Nodes

The Eclair fixes land as operators of Bitcoin’s other major Lightning implementation face a distinct threat. Payment processor BTCPay Server said earlier in September that it observed bots repeatedly probing servers where administrators had manually re-enabled external access to LND.

The attackers targeted an unauthenticated password-change endpoint during the brief window when an LND wallet sits locked, and a successful hit would let them replace the wallet password and request an administrator macaroon controlling the node.

BTCPay responded by issuing unique passwords per LND wallet and blocking unauthenticated wallet-management routes at its network edge, while advising operators against manually exposing the LND API at all.

Together the two incidents point to attackers systematically probing Lightning software for weaknesses that let them seize or redirect funds. As more builders ship Bitcoin-adjacent code quickly, the burden of catching these edge cases before deployment falls increasingly on protocol maintainers rather than automated tooling.

The CCS read. For institutions running or custodying Lightning liquidity, the lesson is operational, not speculative: patch cadence and default fee ceilings now function as risk controls comparable to key management. Firms that let node software drift unpatched are effectively underwriting counterparty risk against every peer they route through, a cost that belongs on the same ledger as custody and slippage.

ACINQ has not disclosed whether any operator lost funds to the closing-fee exploit before the September 14 patch, leaving that question open for node operators reviewing their own historical channel closures against the fixed logic in Eclair 0.14.3.

Get this in your inboxThe Crypto Coin Show newsletter covers the policy and market moves institutional crypto investors are pricing in.

Subscribe

Load More