The potential need for cryptographic upgrades highlights the ongoing challenge of securing digital communications against evolving AI threats.
The post Vitalik Buterin says public-key cryptography may need one or two more upgrades appeared first on Crypto Briefing.
The incident highlights the growing risks of crypto ownership, emphasizing the need for enhanced security measures and regulatory oversight.
The post UK crypto trader robbed of about $800,000 in violent home invasion appeared first on Crypto Briefing.
An AI market correction could significantly impact global economic stability, affecting investment confidence and economic output projections.
The post IMF’s Vamvakidis warns an AI market correction could spill over into the real economy, depending on its severity appeared first on Crypto Briefing.
AI risks could destabilize global finance, necessitating enhanced cross-border cooperation and robust contingency planning to mitigate systemic threats.
The post IMF warns AI risks could push more global banks to fail appeared first on Crypto Briefing.
Valente's optimism suggests Bitcoin's resilience, highlighting the crypto community's proactive stance on evolving security challenges.
The post ARK Invest’s Lorenzo Valente is more optimistic Bitcoin will beat Q Day appeared first on Crypto Briefing.
Bitcoin Magazine

Moscow Exchange to Launch Bitcoin Trading, Testing Underway
The Moscow Exchange plans to let clients trade cryptocurrencies starting in December and has begun testing operations with professional market participants, a senior executive said last week.
“We are actively testing. We started a week ago,” Boris Blokhin, the exchange’s senior managing director for sales and business development, told journalists on the sidelines of the Sber Investments forum, according to a Thursday Interfax report.
The exchange — the biggest in Russia — will work with a partner settlement digital depository, where market participants will open accounts. It won’t need any additional permits or licenses to launch trading, Blokhin added.
The plan follows Russia’s law on digital currencies and digital rights, which took effect in September and sets comprehensive rules for crypto in the country.
Trades can run through existing infrastructure such as exchanges, brokers and trust managers. New players will also join, including crypto exchangers, where assets are bought and sold, and digital depositories, which record ownership.
Russia’s largest bank, Sberbank, announced last month it was the first in the country to have been approved to custody bitcoin. The Bank of Russia gave it the green light for “digital currency record-keeping” and to facilitate the transfer of such assets, covering bitcoin and other, smaller cryptocurrencies.
Russia’s relationship with Bitcoin is complicated. President Vladimir Putin in 2024 seemed to praise the original cryptocurrency back in 2024. “For example, Bitcoin, who can ban it? Nobody,” he said at a forum at the time.
But the Kremlin keeps a tight grip on what citizens can do with it. Using crypto as a form of payment has been illegal in Russia since 2022; unqualified investors can buy bitcoin and stablecoins capped at 300,000 rubles ($3,556) a year through a single intermediary.
Qualified investors can buy any cryptocurrency traded on exchange and over-the-counter markets, with no limit.
President Putin in October signed a law to set in stone the regulation of digital currencies and digital rights in the country.
This post Moscow Exchange to Launch Bitcoin Trading, Testing Underway first appeared on Bitcoin Magazine and is written by Mathew Di Salvo.
Bitcoin Magazine

South Korean Lender Jeonbuk Bank to Offer Bitcoin-Backed Loans
Jeonbuk Bank has become the first bank in South Korea to offer won-denominated loans backed by bitcoin, according to reports.
The subsidiary of JB Financial Group has started offering the loans, according to exclusive Monday reporting by the country’s Edaily paper.
For now, the product is limited to corporate customers raising working capital, and bitcoin is the only eligible collateral. The bank said it may expand the service depending on regulatory changes and market demand, the report added.
Jeonbuk Bank won’t own or control the bitcoin itself. The collateral is held by a specialized custodian, limited to virtual asset service providers registered under Korea’s Act on the Reporting of Specific Financial Information, it further noted.
If a borrower defaults, the bank won’t take the coins but will rather liquidate them through enforcement procedures such as a court-ordered sale and recover the loan from the proceeds.
The bank said it chose bitcoin for its market capitalization and trading volume, the largest of any digital asset, and that its own stress tests showed bitcoin held up as collateral even during sharp price declines. To value the collateral, it uses an index based on a weighted average of prices across multiple global exchanges.
Similar products already exist abroad. JPMorgan Chase debuted bitcoin-backed loans for institutional clients in March, and Switzerland’s Sygnum Bank offers comparable credit.
The loan is the first concrete result of the “Crypto Bank” vision put forward by Jeonbuk Bank President Park Chun-won. The lender is the only regional bank to provide real-name verified accounts to crypto exchanges, and it is preparing an “all-in-one wallet” that combines fiat accounts and digital asset wallets.
“This loan is an example of designing a risk management and recovery structure that allows us to utilize existing bank lending and collateral management systems without directly holding virtual assets,” said Jeong Sang-hoon, executive vice president of the bank’s digital division.
This post South Korean Lender Jeonbuk Bank to Offer Bitcoin-Backed Loans first appeared on Bitcoin Magazine and is written by Mathew Di Salvo.
Bitcoin Magazine

Bitcoin Extends Slide — Can ‘Uptober’ 2026 Beat Last Year’s?
Bitcoin’s price dipped further on Monday, sliding 2% over a 24-hour period to trade at a little over $82,423.
The coin over the past week has dropped by 3%, and so far this month, it’s trading down.
October is usually a strong month for the OG cryptocurrency. Data from Coinglass shows that it has finished eight of the last 10 Octobers in the green.
Last year was one of the exceptions: while bitcoin’s price notched an all-time high of $126,080, it then started to enter a bear market after October 10 — when the biggest crypto liquidation event in history battered the asset’s price.
Markets across the board got hit hard following tariff threats from U.S. President Donald Trump. Crypto long positions got wiped out and over $19 billion in digital asset bets were liquidated.
Bitcoin then spent most of 2026 in a bear market.
The good news? That the coin just had its best quarter in years and a good September — historically a rough month for the cryptocurrency. Analysts now are saying that while it’s got off to a bad start so far in “Uptober”, signs of a bottom this month may already be in.
Writing for data firm CryptoQuant, Japanese asset management platform XWIN said it could be up from here.
“Now, signs of a potential bottom are emerging,” XWIN said in a note. “On October 12, markets priced roughly an 18% chance of an October Fed hike, while Brent crude had eased from nearly $110 to around $103.”
It added: “‘Uptober’ need not mean uninterrupted gains: a correction can build the foundation for the next advance.”
High oil prices have — at least in the short-term — hurt bitcoin’s price as investors then expect the Federal Reserve to hike interest rates in order to control inflation.
But while the U.S. central bank last month raised interest rates, bitcoin’s price was unfazed and rose following the news.
Other analysts have now said the biggest cryptocurrency is out of a bear market due to it crossing its 365-day moving average — a signal that has marked the start of its bull markets in past cycles.
This post Bitcoin Extends Slide — Can ‘Uptober’ 2026 Beat Last Year’s? first appeared on Bitcoin Magazine and is written by Mathew Di Salvo.
Bitcoin Magazine

Ledger Investigates Fund Losses Tied to Southeast Asian Reseller After Implant Found in Device
Ledger has said it is working with authorities after users last week reported losing funds and one received a tampered device. But the company reassured customers that the number of impacted products was “limited.”
The hardware wallet manufacturer added in an announcement on Sunday that it had reached out to affected customers.
Paris-based Ledger on Friday said it had started an investigation after customers in South East Asia reported they’d lost funds after buying from a reseller.
Then, on Saturday, Ledger said it confirmed that one of the impacted users’ devices contained an unauthorized hardware implant.
Ledger would not confirm the amount of money lost but blockchain investigator Specter said $86 million had been drained from devices.
“As the investigation continues, we are working with relevant authorities,” Ledger wrote on Sunday.
It added: “Also, we strongly advise users not to purchase from unauthorized resellers, where the risk of buying suspicious, counterfeit or modified devices is important as nothing operates under ledger security guidelines.
“Security isn’t static and this incident underscores why we continuously review and strengthen our distribution and anti-tamper processes.”
Customers in South East Asia last week reported losing funds after buying from Kuala Lumpur, Malaysia-based vendor CryptoBilis.
Ledger then advised customers who’d bought from CryptoBilis to not set up their devices.
CryptoBilis has since stopped selling its hardware wallets, Ledger said on Sunday.
People on social media posted picture evidence of devices implanted into Ledger products, claiming it could get users’ private keys. It isn’t clear how many of Ledger’s signing devices contained the implant.
Neither Ledger nor CryptoBilis immediately responded to questions from Bitcoin Magazine.
Hardware wallet security has been a hot topic ever since hackers in July were able to steal close to $120 million in bitcoin from Coldcard users.
The products, made by Canadian company Coinkite, contained a firmware bug which led to faulty seed generation, allowing hackers to essentially guess investor seedphrases.
In a separate incident, hardware wallet manufacturer Trezor last month reported that close to 81,000 customers had their details leaked after its third-party fulfillment partner had data stolen.
Criminals have been targeting data this year, with scammers getting hold of customer information via crypto wallet Ledger’s payment processor Global-e to send phishing emails.
This post Ledger Investigates Fund Losses Tied to Southeast Asian Reseller After Implant Found in Device first appeared on Bitcoin Magazine and is written by Mathew Di Salvo.
Bitcoin Magazine

Strategy Smash Buys $21.5M in BTC as It Continues Bitcoin Accumulation
Bitcoin juggernaut Strategy has bought more BTC as it pivots back to accumulating on a weekly basis.
The Nasdaq-listed bitcoin treasury spent $21.5 million on 250 bitcoins from October 5 and October 11, according to a Monday filing.
It added that it also spent $192.4 million repurchasing its STRC preferred stock.
Strategy now holds 848,250 bitcoins worth $70.2 billion at today’s prices. Its stock (MSTR) on Monday was trading nearly 2% higher.
The company — formerly MicroStrategy — has spent the past four weeks buying bitcoins after months of more sporadic buys or sometimes halting purchases for months altogether in order to build a bigger cash buffer.
Monday’s filing also shows that the company bought last week’s bitcoin using net proceeds from MSTR Stock sales with its USD Cash buffer. Strategy has two cash reserves: “USD Reserve”, used to support the payment of dividends on the company’s preferred stock offerings and interest on its outstanding indebtedness, and “USD Cash”, which it uses to buy bitcoin.
It first bought the cryptocurrency in 2020 to protect its shareholders from inflation. Since then, it has aggressively bought the asset and pivoted to being a bitcoin treasury.
Investors can now buy its shares to get heightened exposure to the cryptocurrency, or get paid a yield via its digital credit products.
The company is now focusing on products that pay investors yield via its STRF, STRC, STRK, and STRD preferred shares.
Its Nasdaq-listed stock is down over 50% over the past year after the price of bitcoin took a hit.
But since the company started buying bitcoin in August 2020, MSTR has appreciated by nearly 1,000%.
This post Strategy Smash Buys $21.5M in BTC as It Continues Bitcoin Accumulation first appeared on Bitcoin Magazine and is written by Mathew Di Salvo.
HTTP error 429 on https://cryptoslate.com/feed/
Failed to fetch feed.
Chainlink has issued invitations to New York for 29 October. The event is called Link:NYC, it runs for one day, and the organiser describes it on its own page as a meeting point for executives rebuilding the global financial markets. If you hold Chainlink, you will see an entry in the calendar first of all. What makes it interesting is a number that has nothing to do with the price: 28.
That is how many days it takes for staked LINK to become freely available again. Start the exit today and you reach your coins on 9 November, eleven days after the event. Anyone who wants to react on 29 October with LINK sitting in staking has therefore already made that decision today, without noticing.
Chainlink's event page names five topic blocks: institutional real-world assets, agentic financial applications, interoperability, market infrastructure, and stablecoins and payments. A venue is not given there, nor is an attendance figure. Speaker names sit in a tab of their own that reveals nothing publicly so far. Officially settled, then, are the date, the city and the list of topics.
That is less than the figures circulating in recent days. Several crypto outlets wrote of more than 35 executives expected and listed institutions said to be attending. Chainlink itself confirms none of those details, and the organisers have pointed out explicitly that the presence of individual houses is not an announcement of new partnerships. Anyone treating the 35 as a settled figure is working with a number nobody has evidenced so far. The official description is on Chainlink's event page.
The target audience, by contrast, Chainlink names clearly. Banks, operators of market infrastructure, asset managers and institutional allocators are invited. This group does not decide on code but on budgets, and on the question of which infrastructure a house will settle on two years from now. Link:NYC is explicitly not a developer conference.
That is why dates like this move LINK at all. Chainlink does not sell an application to retail customers. The project supplies data connections and transfer routes between chains, and between chains and the old financial world. Every contract with a large house is therefore news about future revenue, not present revenue. With LINK the market regularly trades the expectation rather than the revenue.

On Monday evening LINK trades at $13.07, or EUR 11.72. That is 1.7 percent less than 24 hours earlier and 6.7 percent less than a week ago. Over 30 days, a gain of 12.6 percent still stands. Market capitalisation is $10.0 billion, 766.8 million LINK are in circulation, and coins worth $295.3 million changed hands in 24 hours. The figures come from CoinGecko, as of Monday 12 October.
The daily range ran from $12.85 to $13.49. That leaves the price some three quarters below the all-time high of $52.70. Between today and 29 October lie 17 days in which none of these figures stands still.
A date on its own says little. More useful is a look at what happened at Chainlink after comparable announcements, and there were two in September.
On 29 September we reported on 11,500 institutions being connected to the Swift ledger, with a quote from Chainlink co-founder Sergey Nazarov. The price jumped above $15 at the time. A day later the counter-move was already in our own pages: LINK slid from $15.44 to $14.14, a drop of 8.4 percent, without anything about the Swift news having changed. Measured against that high, the price today sits 15.4 percent lower.
The pattern is nothing new for LINK. An institutional story carries the price for hours, rarely for weeks, because it confirms an expectation already in the price. For 29 October that means the move, judging by recent weeks' own experience, tends to come before the date rather than after it. That is not proven; it is an observation from two cases.
Alongside the price, a movement is running in the wallets that a chart does not show. Addresses holding between one million and 100 million LINK held 65.6 percent of the circulating supply on 12 October. On 1 October it was 65.38 percent. The change is small but the direction is unambiguous: while the price gave way over that period, coins moved into larger addresses. The survey comes from TokenPost and names no data provider, which is why it serves as an indication and not as proof.
Two points matter about it. High concentration means comparatively few addresses can move the market. And part of those holdings sits in staking, so it cannot be sold immediately, which cushions downward moves and delays upward ones.
One point has been going missing in the coverage of Link:NYC so far. The Federal Reserve's meeting calendar lists a session of the open market committee for 27 and 28 October, with the decision falling on 28 October. Link:NYC lies exactly one day after it.
For the market that means two things. On Wednesday evening Central European time, a rate decision moves the entire crypto market, Chainlink included, regardless of what Chainlink itself does. On Thursday, news out of New York meets a market still digesting the rate move. Anyone attributing the price reaction on 29 October to the event may be confusing two causes. Know that sequence before you read a move on the evening of the 29th.
This is where the date turns into an action. Chainlink staking has run since version 0.2 with a cooldown of 28 days followed by a withdrawal window of seven days. Both are set out in the official overview of Chainlink Staking v0.2.
Work it through on your own calendar. Start the exit today, 12 October, and the cooldown ends on 9 November. From then until 16 November there is time to withdraw the LINK. Miss that window and the coins move automatically back into staking, with the 28 days starting over. On 29 October, the day of the event, you are in the middle of the period in every one of those cases.
A plain consequence follows. On 29 October, staked LINK is not a holding you can dispose of but one you watch. Anyone who wants it otherwise acts today and forgoes the staking income of the next four weeks. Which providers set which periods and rates is in the comparison of staking providers; the periods differ considerably between staking in the protocol and staking through an exchange.
A second deadline runs independently of that. Buy LINK today and hold the coins as private assets, and the one-year speculation period under Section 23 of the Income Tax Act ends on 12 October 2027. From 13 October 2027, a gain on sale stays tax free. That applies to real coins, not to a certificate on LINK; we worked that distinction through on 10 October using the LINK funds as an example.

On days with heavy trading volume, the result depends less on the choice of exchange than on whether you have access at the right moment. Three points can be checked beforehand, without haste.
If your LINK sits on an exchange, check whether the withdrawal function is enabled and whether a change to two-factor authentication is still pending. Many houses lock a newly stored withdrawal address for 24 to 72 hours. Anyone seeing that lock for the first time on 29 October has seen it too late. Which providers in Germany are licensed under the European crypto regulation MiCA, and what their fees look like, is in the exchange comparison.
If the holding sits on a hardware wallet, check the firmware before the date and whether the recovery words are where you can reach them if it matters. A device that demands an update first costs half an hour at the decisive moment. Check too whether your device supports the token at all before you send coins to it.
Note the purchase date, amount and price as soon as you trade. For the one-year period the day counts, not the month, and the tax office accepts no estimate. An export from the exchange on the day of purchase saves a reconstruction the following year.
Two values from current trading can be checked and serve as orientation. Above sits the daily high at $13.49. It lies around 3.2 percent above the current price and marks the point at which the market last sold today. Below, the daily low sits at $12.85, a good 1.7 percent under the price.
Further down waits the weekly low of 8 October at $12.09, which TokenPost reports for last Thursday. Between that low and today's price lie 8.1 percent. As long as the price stays above it, the move since Thursday is intact; fall below and the recovery of the past four days has been worked off. Both levels come from trading and not from a model, and neither is a forecast.
From this newsroom's point of view, Link:NYC is for the next two and a half weeks above all a date that expectations can attach to, and not yet an event with documented content. Three observations argue for that. Chainlink has announced neither speakers nor outcomes. The attendance figures in circulation are second-hand. And the two institutional stories in September, Swift on the 29th and the Infosys connection the week before, each carried the price for hours and then gave way; LINK today sits 15.4 percent off the high of 30 September.
Against that stands the timing. A meeting with banks and market infrastructure one day after a rate decision can amplify a move that is running anyway, and the shift into larger addresses shows that somebody is accumulating at today's prices. Anyone drawing a purchase decision from that is betting on an expectation and not on a number. A total loss is possible at any time with crypto assets.
Three steps can be worked through before the date, and all three cost nothing.
(As of October 12, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
Cardano processed exactly 1,780,896 transactions in the five days between 6 and 11 October 2026. The chain has never moved more in its history; the previous high, from November 2021, sits 13.7 percent below it. The Cardano price has not responded: ADA stands at around $0.2384, about 5.3 percent below where it was 24 hours earlier.
Most reports name the Cardano governance votes as the cause, above all the proposal for a minimum fee of 75 ADA for stake pools. That explanation does not add up arithmetically. We counted the votes on that proposal one by one: there were 148. This article sets out what the chain data supports about the record, why the proposal has since lapsed, and what follows from it if you hold ADA in Germany.
The figures in this article rest on the public block data of the Cardano chain itself, analysed across all 662 epochs since the chain launched, and on the full list of votes on the proposal in question. cryptoticker.io compiled this analysis itself on 12 October 2026.
Cardano counts in epochs rather than days. An epoch is a fixed stretch of five days in which the chain produces blocks, settles rewards and fixes voting rights. Epoch 660 began on 6 October 2026 at 9:44 p.m. and ended on 11 October at the same time.
Within that window, 1,780,896 transactions were written into 21,243 blocks. For comparison: epoch 659, the five days before, carried 158,011 transactions. The jump is 11.3-fold. Epochs 653 to 657 all sat between 89,353 and 130,442 transactions, in other words within the usual band. Epoch 660 is not a trend but a single, extremely sharp spike.
The old record stood at 1,566,974 transactions and was set in epoch 304, which ran from 21 to 26 November 2021. ADA was then just past its all-time high, the market was in the closing phase of the last big cycle, and the chain had only recently gained smart contracts.
That comparison is the genuinely interesting part, and most coverage leaves it out: in 2021 the load came out of a hot market with prices to match. In 2026 it comes at a price 92.3 percent below the all-time high of $3.09. The same throughput, an entirely different price base. Anyone reading network activity as a price driver gets a counter-example here straight from the chain data.
The decisive observation lies in the ratio, not in the total. Cardano produces a quantity of blocks per epoch set by the chain, and that quantity has not changed: 21,243 blocks in epoch 660 against 21,207 in epoch 659. A difference of 36 blocks, which is to say nothing.
Per block that works out at 83.8 transactions against 7.45 the week before. The blocks were roughly eleven times fuller, not more frequent. The look back is interesting too: in the 2021 record epoch 304, a block held 74.3 transactions at an almost identical block count. The chain packed more into the same frame in 2026 than in 2021, even though block size has not multiplied in that time.
For the technology that means the load was not spread across a larger infrastructure, it was pressed into the existing one. The chain carried that without an outage, and that is the solid finding of the week. It has nothing to do with a capacity upgrade or a protocol change.

The fees draw the same picture from a second angle. Epoch 660 generated 337,720 ADA in transaction fees, where epoch 659 produced 49,495 ADA. The factor is 6.8.
Notably, fees rose less steeply than the transaction count, 6.8-fold against 11.3-fold. In arithmetic terms that means an average fee of 0.19 ADA per transaction in epoch 660, against 0.31 ADA in epoch 659. The transactions were more simply built than usual, then: few inputs and outputs, no demanding smart-contract calls. On Cardano the fee follows the size and complexity of a transaction, not an auction procedure as on Ethereum. A mass event made up of many small, uniform transfers therefore produces exactly this pattern.
At 337,720 ADA and a price around $0.2384, we are talking about roughly $80,500 in fees over five days. For the stake pool operators, whose remuneration was at issue over the same period, that is not a sum that turns a situation around. Precisely that is why the fee question became political on Cardano at all.
minPoolCost is the fixed amount a stake pool retains from rewards each epoch before anything at all is distributed to delegators. That amount stands at 170 ADA today. The proposal voted on since epoch 654 wanted to cut it to 75 ADA. Small pools would have become more competitive, and delegator returns would have shifted slightly.
When we reported on this vote on the minimum fee on 10 October, it stood at 34 percent in favour with the deadline approaching on 11 October. The result is now in, and it moved, but not enough: the delegated representatives' yes votes ended at 44.18 percent of the active voting power, the no votes at 55.82 percent. The majority was therefore missed. The proposal lapsed when its deadline passed in epoch 661, meaning it was neither adopted nor enacted. The 100,000 ADA deposit goes back to the submitters, and the minimum fee stays at 170 ADA.
For delegators that means, concretely: nothing changes in how your staking rewards are distributed. How hard that fixed amount actually presses on returns depends on the size of your pool; across the pools we recalculated at the start of October, returns ran between 2.1 and 2.7 percent a year.
Now to the question of cause. Several trade outlets trace the transaction record to a rush on the governance votes: users are said to have moved their wallets in order to vote on network parameters, with the deadline coinciding with the end of the epoch. That sounds plausible, because the timings line up.
It can be counted, though, and we counted it. The full list of votes on the minPoolCost proposal contains 148 entries: 87 yes votes and 37 no votes from delegated representatives, plus 19 abstentions and 5 yes votes from the constitutional committee. Each of those votes is one transaction on the chain.
148 transactions against 1,780,896. That is 0.008 percent of the record. Even counting every other open proposal in the epoch and assuming similar participation for each, you do not get beyond a few hundred votes. The governance explanation does not carry the record even as an order of magnitude.
The reason lies in how the system is built: on Cardano it is not individual holders who vote but DReps, delegated representatives who exercise the bundled voting weight of many wallets. A holder who has delegated their voting power to a DRep once need not send a single transaction for any further vote. The whole purpose of the design lies in that bundling. A voting deadline on Cardano therefore produces almost no chain load, structurally.
The vote list shows something else that matters more for placing the vote itself than for the record: not one stake pool voted actively on this proposal. 629 pools sit on the permanent default of "always abstain", bringing a voting weight of some 12.22 billion ADA with them, and 17 more on "no confidence". Among the delegated representatives, a further 10.5 billion ADA or so sits on permanent abstention.
Measured against the active stake of roughly 21.27 billion ADA in this epoch, Cardano parameters are therefore decided by a fairly small, alert share of participants, while the large remainder parks on a default setting. That distribution is what the chain data says and is no judgement on those involved. Anyone who does not want to leave their voting weight unused has to delegate it actively to a representative; with a hardware wallet in self-custody that works directly from the wallet, while with ADA on an exchange it generally does not work at all.

If it was not the vote, then what? The only concrete attribution so far comes from the epoch report of a stake pool operator, quoted by the outlet TechBullion. On that account, some 1.08 million transactions fell in the first three days of epoch 660 alone, and most of them traced back to a campaign around the Cardano token HOSKY, in which participants were to send ADA to an address and receive tokens back. A post on the exchange platform KuCoin credits the same campaign with a peak of 440 transactions per second in a single block.
We cannot confirm that attribution. Both accounts rest on the same source, and we have not recalculated a single campaign's share of 1.78 million transactions from the chain data. What we can say is that the pattern fits. An event in which many people send small amounts back and forth produces exactly what the data shows, namely very many very simple transactions at a below-average fee each. The vote does not produce that pattern.
Our assessment: the record is real and the chain carried it without disruption, but it is not a demand signal for ADA. Three pieces of evidence point that way. First, 148 votes do not explain 1.78 million transactions. Second, the fee per transaction, at 0.19 ADA, came in below the previous week's, which argues for mass simple transactions and against economically heavy use. Third, the current epoch 661 is already well lower again. Against that it could be objected that even an artificially generated load test proves resilience, and that is true. Only nobody pays a higher ADA price for resilience.
A second angle, alongside price and transaction count, is the capital actually at work on the chain. Around $66.5 million is currently deposited in Cardano applications. That means the sum working in trading venues, lending protocols and similar applications on the chain.
That figure puts the record in proportion. Cardano carries a market capitalisation of roughly $8.95 billion, which places it 17th among the largest cryptocurrencies; 37.55 of a maximum 45 billion ADA are in circulation. Against that stand $66.5 million of active capital in the applications. For comparison: on Solana that value is around $6.1 billion, on NEAR around $262 million. A record number of transactions coming off such a narrow base of capital points to many small movements rather than to new money.
The price itself gives the record week no confirmation. ADA trades at around $0.2384, after a daily high of $0.2556. Over 24 hours that is about 5.3 percent lost, over seven days around 11.4 percent. Only over 30 days does a gain of about 14.8 percent stand. The all-time high of $3.09 from September 2021 sits 92.3 percent above today's level. Trading volume over the past 24 hours comes to roughly $376 million.
The pattern is striking: the most transaction-heavy week in the chain's history coincides with a week in which ADA loses a good tenth of its value. Anyone reading the record as a reason to buy should read that sentence twice. Network activity and price ran in opposite directions on Cardano this week, and the chain data supplies an explanation for it in the fee per transaction.
A failed vote and a throughput record yield four concrete points for you.
The minimum fee stays at 170 ADA. Do not calculate your staking income with 75 ADA. The fixed amount is retained per epoch and per pool before anything is distributed. With a small pool it eats a markedly larger share of your reward than with a large one. Anyone who meant to move their delegation depending on the outcome of the vote now has no occasion to.
Your voting weight does not lapse, it rests. As long as you have not delegated your voting power to a representative, it counts as an abstention in every vote. That is the default setting where, according to the chain data, the largest share of the stake sits. Delegating and staking are two separate acts: you can have your ADA in a pool and your voting weight nowhere at all.
Staking income is taxable in Germany. Rewards from staking count as other income and are taxed in the year they are received. For the ADA itself, the one-year holding period for private assets applies: sell at a gain within a year and that gain is taxable, after that it is not. Whether and how staking affects that period depends on the arrangement; anyone delegating regularly should document the inflows as they come rather than reconstruct them in March of the following year.
The purchase route decides your access to governance. If your ADA sits on an exchange, the exchange holds the keys. You generally cannot vote then, and delegating to a representative drops away too. Anyone who wants that needs the coins in their own wallet. Anyone who does not need it gets along more easily with an exchange licensed in the EU.
As of this analysis, the current epoch 661 has processed 118,374 transactions in 4,269 blocks. That is 27.7 transactions per block: clearly above the normal band of around 7 in the preceding weeks, but far below the 83.8 of the record epoch. The spike is not holding, in other words. Three steps follow from that:
Both can be checked without a detour: the figures for epoch 660 are on Cexplorer's epoch page, and the outcomes of the votes are in Cardano's official governance portal.
(As of October 12, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
If you hold G tokens, USDC, USDT, WETH, DAI or WBTC on Gravity Alpha Mainnet, you have until 1 November 2026 to get them off the blockchain. Galxe is shutting down its first own network that day. Anything still sitting on it after that can no longer be transferred, the provider says: not a temporary halt, but the end of mobility. Galxe's own guide therefore explicitly advises bridging balances out before the date.
Two things make that deadline harder than it looks on paper. First, different tokens use different exits, and they are not interchangeable: anyone trying to send USDC over the familiar bridge will not find the balance there at all. Second, the classic route carries a waiting period of roughly seven days. Together that means the real edge in the calendar is not 1 November but a week earlier.
Gravity Alpha Mainnet is the older version of the Gravity network and ran as a layer 2, a side chain that draws its security from Ethereum. Galxe launched it in 2024 on the Arbitrum Nitro stack, with chain ID 1625. With Gravity L1, a standalone layer 1 with ID 127001, now in full operation, the Alpha chain has lost its purpose.
The developer documentation now lists the old chain openly as "legacy" and puts a warning in front of it: all balances should be moved back to Ethereum before 1 November. The help pages for end users put the same sentence more plainly. Gravity Alpha Mainnet will remain available until 1 November 2026, it says, after which it will be decommissioned, and whatever is still on it will no longer be transferable.
For holders this is not a price question. G traded at around $0.0041 on Monday evening, its market capitalisation stood at just under $30 million, which places it 524th among crypto assets. The price sits about 82 percent below its February 2025 peak of roughly $0.023. The sums are small for most holders, in other words. The difference between a small amount and a lost amount is the same all the same, and that is the point here.
Galxe announced the step back in July. The reasoning: Gravity Alpha was the first stage of its own blockchain plans, carried real traffic and exposed the bottlenecks that mattered when building the new chain. Out of that experience, by the company's account, came parallel execution of smart contracts, confirmation times below one second and a price oracle that sits directly in the protocol. An oracle is the feed through which a blockchain learns data from outside, such as prices.
Put differently, the Alpha chain was a test rig that has served its purpose. That fits a development we have described several times this year. Blast is being discontinued, Abstract Chain faced the same question in October, Mintchain also set a deadline, and at NEAR it was MyNearWallet that went rather than the chain. The reason is the same in each case: too many side chains compete for too little usage, and whoever runs one pays for every block.
This is where most of the mistakes happen. Two different kinds of token sit on Gravity Alpha, and they leave the chain by separate routes. What decides is how a token got there in the first place.
Only three assets come out through the official Gravity bridge: the network's own G, plus DAI and WBTC that were deposited through the standard Arbitrum gateway. For those three the canonical bridge, the exit built into the rollup, is the one that applies.
USDC, USDT and ETH or WETH do not come out there. Those three were issued on Gravity Alpha by Stargate, a service that works with liquidity pools, and not by the standard gateway. The canonical bridge does not know them and cannot process them. The Gravity documentation flags this as a hazard warning and adds a sentence worth taking seriously: no further route will be added. For these three tokens, Stargate is the only exit.
One detail that will otherwise send you down the wrong path: on Stargate, USDC is not called USDC but USDC.e. Search for "USDC" and you may not find your balance and take it for gone. The three assets appear there as USDC.e, USDT and WETH with Gravity as the source network.

Gravity Alpha was an optimistic rollup. The name describes the procedure: the chain first assumes its settlements are correct, then leaves others a window to object. That window is called the challenge period and runs for roughly seven days on Gravity Alpha. Only after it can you collect a withdrawal on Ethereum.
The documentation therefore says to start well before 1 November so the withdrawal can complete. Do the arithmetic: start on 25 October and the standard route, with its seven-day wait, just fits inside the window. Start on 29 October and that route no longer gets you out.
One more point the help pages set out separately: withdrawals you have already submitted through the canonical bridge continue under the old procedure and cannot be sped up after the fact. A transfer already running on a seven-day clock stays on seven days, even if you hear about the faster route afterwards.
For G there is now a second exit, and Galxe itself calls it the recommended choice. Fast Withdraw sits on the official Gravity bridge and sends G to Ethereum without the seven-day wait. The guide gives about ten minutes as the typical duration.
The process has four stations: open the bridge, connect the wallet, select Fast Withdraw, enter the amount and confirm the withdrawal in the wallet. Once the G tokens have arrived on Ethereum, set the destination to Mainnet and bridge on from Ethereum to Gravity L1. For that second step you need some ETH in the wallet, because the network fee on Ethereum is paid in ETH. That is the item holders of small G balances should price in beforehand: the fee depends on how busy Ethereum is, not on the size of your balance.
Anyone who needs G on Ethereum faster and would rather not take the official route will find a pointer in the documentation to Symbiosis, a liquidity-pool route that works in minutes. The same caution applies as with any pool service: you trade waiting time for trust in one more provider, and you usually pay a discount for it.
This part is easy to miss and costs the most if you do. The G staking page at Galxe goes with the chain. Staked G cannot be withdrawn afterwards, because the interface for it no longer exists. Anyone with G in staking therefore needs two steps in this order: unstake first, then bridge.
Allow extra time for it. Many staking contracts have a cooldown phase in which a withdrawal is announced but not yet paid out. Check in the interface which period your stake carries before you count on the bridging time. Whether a platform states its withdrawal periods clearly is one of the points our staking comparison looks at, and this case shows why that is no side issue.
If neither the bridge nor the pool route appeals, Galxe's guide names a third way: an exchange that accepts G deposits directly from Gravity Alpha. OKX and Bitget are named there. The process consists of three stations. You send G from Gravity Alpha to your exchange account, withdraw from the exchange over the Ethereum network to your own wallet, and bridge from there to Gravity L1. Here too you need ETH for the fee on the last step.
The route has one solid advantage: you avoid the waiting period. It also has two drawbacks worth knowing in advance. The deposit address has to support the Alpha chain, and it has to still support it when you send. Exchanges often close deposits for a chain before it is switched off. So check the network selection in the deposit dialogue immediately before sending, not against a guide from the week before last. And as long as the tokens sit in the exchange account, the exchange holds the keys, not you: leave them there only as long as the move takes.
The deadline is not the only reason to act now. The contracts for Galxe Quests, meaning Loyalty Points and Token Rewards, have already moved from Gravity Alpha to Gravity L1. Anyone collecting points or rewards with a direct web3 wallet needs G on Gravity L1. The system looks there and nowhere else.
If your G is still on the Alpha chain, the claim fails. The help pages give the exact error message users then see: too few G tokens, even though the balance is visibly there. The fix is the same as above. Check in the wallet which chain your G tokens are actually on, bridge them to Gravity L1, switch the network in the wallet and try the claim again.
Galxe also names a more convenient variant called Smart Balance, which lets network fees across several networks be paid out of one balance. For the plain task of rescuing holdings before 1 November you do not need it; for regular participation in quests it takes work off your hands. If you pick up tokens through tasks like these, our airdrop overview lists the current opportunities.

For context, one point that stands out in the research and that you could stumble over if you read up yourself. The announcement and the exchange notices give 1 November 2026 as the day network operation ends. Gravity's developer documentation gives the same withdrawal date but writes alongside it that the network will be "unsettled" in December 2026, meaning no longer settled.
Do not read that as an extension. Both sources require withdrawal by 1 November, and the December reference concerns the technical wind-down of the chain, not your window. Anyone relying on December is relying on a reading that neither source offers. What binds you is 1 November, and on the standard route 25 October.
For developers more than a balance hangs on the date. Galxe requires that dApps, bots and scripts running on Gravity Alpha be redeployed to Gravity L1. That includes things that run invisibly day to day: a trading bot with a hard-coded chain ID, a monitoring script, a data collector that queries the old endpoint.
Two notes from the documentation help with the move. The chain ID changes from 1625 to 127001, and the contract addresses for G and the other tokens differ between the two networks. Anyone with addresses fixed in code has to swap them, not merely redirect the endpoint. The reference pages for the old chain will remain as an archive, according to Gravity, but the endpoints themselves will not.
A chain switch is no trivial event for tax, and the position here is less clear-cut than it should be. Private disposals of crypto assets in Germany fall under Section 23 of the Income Tax Act: after a holding period of more than a year a gain is tax free, while within the year an exemption threshold of EUR 1,000 applies to all private disposals in a year taken together. Exceed that threshold and the entire gain is taxable, not only the part above it.
What is open is how the individual routes should be classified. When the canonical bridge moves the same token across chains, there is economically no exchange but a transfer. A route over liquidity pools, by contrast, gives you units out of a pool at the destination and can thereby take on the character of a swap, and a swap is a disposal under Section 23. The tax authorities have not commented explicitly on bridging, and a binding classification is not something this text can give you.
What you can do yourself is useful regardless: record the time, the amount and the addresses involved for every step while the old chain still answers queries. After the shutdown you will reconstruct that evidence less easily, because the public endpoints disappear. For larger amounts, or for a pool route, asking a tax adviser is worth the money.
The situation is manageable as long as you deal with it now. Three steps, in this order:
The sources for this article are Galxe's bridging guide and the withdrawal page of the Gravity documentation, which sets out the routes per token and the waiting period in detail.
(As of October 12, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
Ledger has confirmed that at least one device belonging to an affected user contained an unauthorised hardware implant. According to the manufacturer, every confirmed case traces back to a single reseller in South-East Asia. Since 11 October, users have also been reporting devices bought from European retail, and that is where the case starts to matter for you: if the sales channel is the point of attack, the brand on the box counts for less than the question of whose hands your device passed through.
In a status report 72 hours after the incident, Ledger wrote on X that all confirmed cases of the theft disclosed on 9 October involved devices sold through the South-East Asian reseller CryptoBilis. The security publication Help Net Security documents that wording. The manufacturer also confirmed that one affected device contained an unauthorised additional circuit board.
Two things Ledger did not do. It did not explain how the implant works technically. And it did not give the number of affected devices. The company does state that its own systems and its direct sales channel were not compromised. Ledger says it is working with the relevant authorities and is contacting affected users.
That distinction is no side note. An attack on manufacturing would affect every device in a production run. An attack on one station in the distribution chain affects the devices that passed through that station. The difference decides how many people have to act, and so far the evidence points to the second case. How the incident began, we set out on 10 October in our review of the CryptoBilis case.
CryptoBilis is an authorised reseller and serves customers in Indonesia, Malaysia and the Philippines. The company had changed owner shortly before. Ledger told the retailer to stop sales and shipping; on the manufacturer's account, CryptoBilis has ceased selling hardware wallets.
The instruction Ledger then issued to all authorised retailers is the notable part: source goods only through authorised channels, and never return a returned product to the shelf. The second half describes the likely lever. A returned device, opened and resealed, looks like new stock in a shop.
A supply-chain attack does not hit the product at the manufacturer. It hits a station on the way to the buyer: warehouse, transport, returns processing, shop shelf. For buyers, that means the security of a hardware wallet does not begin when the device is first switched on.

The most detailed technical description so far comes not from the manufacturer but from Mark Karpelès, the former head of the Mt. Gox exchange. He says he bought a tampered device himself in Malaysia and has published photographs and video. They show a small circuit board under the screen, in the place that otherwise holds only a piece of padding.
On his account, that board carries a microcontroller, an antenna, an LTE module and an e-SIM. An e-SIM is a permanently fitted, digitally writable SIM card; it needs no card slot and is invisible from the outside. Karpelès describes the component reading along with whatever the screen displays, capturing the recovery phrase in the process and sending it to the attacker over the mobile network.
The recovery phrase, also known as the seed phrase, is the sequence of 24 words from which every private key in your wallet can be recomputed. Anyone who knows it no longer needs your device. That makes setup, the moment when those words appear on the display once, the most valuable moment for an attacker. Ledger has not confirmed this account of how the implant works, and it rests so far on a single investigation, if a technically documented one.
Karpelès also reports a second suspect device, obtained through a listing on Amazon Japan with stock that came from Malaysia. That too points to the trade route rather than to manufacturing.
The most uncomfortable finding of the case concerns the very check buyers have relied on until now. The Genuine Check is the authenticity test in which Ledger's software runs a cryptographic key exchange with the secure chip inside the device and then confirms that the unit is genuine.
Karpelès says his tampered device arrived in intact shrink wrap and passed that test. The reason is obvious once you have heard it: the secure chip was real. The implant sits next to it and the check does not cover it. On his assessment, opening the device is the only reliable way to find the additional board.
There is an awkward consequence for everyday practice. A passed authenticity check is still a good sign, but it is not proof that nothing foreign sits inside the case. Anyone holding a second-hand device, or one from a channel with an unclear chain of custody, cannot rest on that signal alone. Which devices are built in what way, and where the differences lie, our hardware wallet comparison sets out.
There is no official damage figure. The on-chain analysts at Bitquery arrive at a suspected total of more than $93 million. Other independent assessments come out lower; the range of published estimates runs from around $72 million to more than $93 million, spread across hundreds of wallets. One analysis names $93.4 million across more than 300 addresses.
Balances on several blockchains are affected, with Bitcoin, Ethereum and Tron named. The figures diverge because of method: anyone attributing addresses to an incident has to decide which movements still belong to it. We therefore give the range and do not smooth the values into a single number.
The single finding that matters most to you is not the damage total. It sits in a time window. Bitquery traced the drained wallets back and finds that more than eight in ten first received a balance in June or later. That suggests devices bought at CryptoBilis as early as June can be affected too.
That reaches beyond Ledger's own advice. The manufacturer had told buyers from the past 90 days to move their holdings to a new device. Counting back from the incident on 9 October, that window ends in early July. The on-chain trail begins a month earlier. Anyone who bought a device through this retailer in June falls outside the manufacturer's recommendation, but not outside the data.

On 11 October a report from Europe became public. A user writing on X as @0xJohannes_ says he bought his device in a branch of an official European reseller; the channel Crypto Banter carried the case further the same day. On his description, the device held additional cables that do not appear in genuine units, and the build resembled an older variant of the implant even though the purchase was only weeks old. Karpelès, who posts on X as @MagicalTux, pointed to an antenna whose marking suggested a "Eurasian" version.
Care matters more than speed here. No balance was taken from this device. Which retail chain is meant does not emerge from the report itself. It is named only in adjacent coverage; the trade publication Crowdfund Insider writes exactly that and refrains from confirming the case itself. Ledger has not commented in detail on the European report and continues to attribute the confirmed cases to the South-East Asian retailer. Several observers regard the published images as insufficient evidence and warn against inferring broad contamination of European retail from individual cases.
Something is left for your decision all the same. There is one confirmed implant case, a documented route through the intermediate trade, and a manufacturer's instruction not to restock returns. That is enough to retrace your own purchase route once, and not enough to panic.
Ledger's recommendations for affected customers are clearly worded. Anyone who bought at CryptoBilis and has not yet set the device up should not initialise it. Anyone who has already set it up should move their holdings to a new device with a newly generated recovery phrase, ideally one bought directly from the manufacturer. Contact should run through the official channels, and the company explicitly recommends filing a report with the local police.
For the German market there is no confirmed case. The reports from European retail do give reason, though, to make the purchase route the criterion. Buying a device directly from the manufacturer, or from a retailer with a traceable supply chain, costs nothing but a little patience. A device from a marketplace, from a classified ad or as a return from unclear hands carries a risk that cannot be checked from the outside.
What you should not do matters as well: opening your own device to look for a board destroys the warranty and, without reference images and measuring equipment, gives you no reliable answer. The route via a new recovery phrase on a cleanly sourced device is cheaper and safer than any tinkering with the case.
This incident touches a decision many investors take once and then never revisit: self-custody or custody by a service provider. Self-custody means that you hold the private key yourself and nobody else can dispose of your assets.
The price is that there is no counterparty when something goes wrong. Anyone who loses a seed through a tampered device has no claim against a custodian. What remains is a damages claim against whoever tampered with the device, and that person is usually unknown. With a custodian authorised under the EU's MiCA regulation, the customer by contrast holds a claim against a supervised company, with duties to segregate client holdings. In exchange, that customer gives up the key. Which providers hold an EU licence, our overview of regulated crypto exchanges shows.
Both are defensible, and most investors fare better with a split than with a pure solution. What counts is knowing which part of your assets you answer for yourself if the worst happens.
Every major security incident is followed by a second wave, and it reaches more people than the incident itself. Ledger explicitly warns that fraudsters are exploiting the situation and makes one point clear: the company does not call, does not send direct messages and never asks for the 24-word phrase. Any message that does precisely that is an attack, however genuine the sender and the design may look.
The same incident has already pushed fake Ledger pages into search results. How to spot them, we set out in our analysis of the fake Ledger websites. The rule behind it is short: the phrase never leaves the device, on no channel, for no support agent.
In the market the news has moved little so far. Bitcoin trades at around $82,400 on CoinGecko data, a good one percent below the previous day, and the market as a whole lost about 3.7 percent in 24 hours. The incident is grave for those affected and, for the price, a side issue so far.
If a doubt about the purchase route remains, no test on the case will settle it. Only a fresh key on a device whose origin you know. These three steps get you there:
(As of October 12, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
An account on X claimed on Monday that the SHIB airdrop was back, and circulated a link for checking your own wallet. Mazrael, a long-standing member of the Shiba Inu community and an adviser to the game Shiba Eternity, called the post a scam the same day: “care. no such airdrop. this is a drainer.” Anyone holding Shiba Inu loses nothing from a click alone. The holding goes only with a signature, and that is exactly what such pages present as a harmless check.
The fake account posed as an official source and spoke of a new round of free tokens. Anyone following the shortened link landed on a page that at first asked only for the wallet to be connected. A window then opened that looked like an ordinary transaction. Mazrael wrote that such posts were flooding the timelines right now, and advised against clicking and in favour of reporting the accounts. The industry outlet U.Today picked up the episode at 11:25 UTC on October 12.
Two things are missing from the report, and both matter for the assessment. Neither the domain of the scam page nor a loss figure is named. There is therefore no documented number for victims or tokens drained, and anyone reading such a number should ask for its source. What is documented is the episode itself and the warning from someone close to the project.
The project's official point of contact is shib.io. Anything that would be a genuine project airdrop would appear there and in the channels linked from it. An offer that surfaces only on X and presses for haste is therefore suspect before any technical check.
A wallet drainer is not a virus that sneaks onto your computer. It is a contract on the blockchain that the victim itself gives permission to move its tokens. In the language of the Ethereum standards, that permission is called an approval or allowance. It is granted with the approve call, and three details are fixed in the process: which token, which recipient and which amount.
The catch lies in the amount. Many interfaces enter the highest possible value there by default, so the user does not have to sign again on every trade. An unlimited approval of that kind stays in place until someone revokes it. It covers not only today's holding but every SHIB that later arrives at the same address. The attacker does not even have to be in a hurry after that. They can wait weeks and pull the holding out once it has grown large enough.
There are two routes to signing, and the wallet displays them with differing clarity. A transaction costs a fee and appears later in the block explorer. A pure signature under the EIP-2612 standard, known in practice as permit, costs nothing and never shows up in your own history at all. The attacker submits it and pays the fee. Anyone looking only at the transaction list therefore sees nothing unusual, even though the permission was granted long ago.
Here is how to recognise the situation: when a page promises free tokens but demands a signature granting the other side access to your holding, the direction is wrong. A genuine airdrop sends tokens to your address. It needs no permission from you, because nobody requires authorisation to send you something.

Memecoins suit this trick especially well for two reasons. The first is the unit count. In circulation, according to CoinGecko data, are around 589.24 trillion SHIB, with total supply at 589.50 trillion. Of the original quadrillion, 41.05 percent is therefore gone, burned or permanently locked. Anyone seeing quantities like that in their portfolio easily loses the feel for the value behind them. Ten million SHIB sounds like a lot and is worth around 47.80 euros on Monday evening.
The second reason is expectation. Shiba Inu grew out of an airdrop and out of distribution campaigns, and that is precisely the story the fraudsters retell. Add to that the state of the price: after a year down 49.9 percent, the hope of free tokens is an effective lever. The longer a holding sits at a loss, the more readily its owner reaches for an offer that evens out the sum.
On Monday evening SHIB trades at $0.00000533, or 0.00000478 euros, according to CoinGecko. The daily range runs from $0.00000530 to $0.00000555, and the 24-hour loss comes to 2.6 percent. Over the week there is a 9.3 percent loss, over 30 days a gain of 0.6 percent, and over one year a loss of 49.9 percent. Market capitalisation stands at $3.14 billion, or 2.81 billion euros, ranking 37th. In trading, $63.1 million changed hands within a day.
SHIB is 93.8 percent short of its all-time high of $0.00008616 from October 27, 2021. For the scam, that figure is no footnote: bridging exactly that gap is what an offer of free tokens promises.
Yesterday, in this spot, we worked through Shibarium's network utilisation after the 151 percent jump. The price has moved a further 2.6 percent down since then, and the daily low of $0.00000530 sits close to the current level. The mark that still left room below it yesterday has become the day's line. Nothing has changed on the chain itself in these 24 hours that would explain the price. The pressure comes from the broad market, in which Bitcoin and Ethereum are also down over the week.
The trick works only where you sign yourself. If your SHIB sits with an exchange, no page on the internet can get you to grant an approval, because you do not hold the keys at all. In exchange you carry the custodian's risk. Since December 30, 2024, providers in the EU need authorisation under the MiCA regulation for this business, and a scam page never holds that authorisation. Deposit protection of the kind that applies to a bank account does not exist for crypto, not even with a licensed provider.
If you hold your own keys, the relationship reverses. Nobody can freeze your holding, but every signature is yours. A hardware wallet shifts that line once more: the key never leaves the device, and the operation has to be confirmed on the device. Against an approval you confirm on the device yourself, however, that does not help. It protects against theft of the key, not against your own consent. If you want to compare the devices and their differences, you will find them in our hardware wallet comparison.
A middle route that stands up well to this trick is separation by purpose. One address for day-to-day use with small amounts, a second address without any approval at all for the holding that is meant to sit still. Connect only the first to outside pages and the damage stays capped, in the worst case, at the small amount.

Revocation is the only operation that undoes an approval once granted. Technically it sets the permitted amount to zero. It costs a network fee, because it is itself a transaction, and it takes effect as soon as it is confirmed in a block.
Here is how to go about it. Open an approval checker such as revoke.cash or the corresponding page of a block explorer and enter your address. The list shows, for each token, which contract may move how much. Pay particular attention to entries with unlimited amounts and to contracts whose names you cannot place. Set each of these approvals to zero, starting with the largest position.
Two points are often overlooked. First, the list always covers one chain only. SHIB runs on Ethereum, on Shibarium and, since the start of October, in bridged form on Solana as well, and each chain keeps its own approvals. Second, a signature granted under EIP-2612 does not disappear from the list if it has not yet been submitted. If you suspect such a signature was handed out, the only remedy left is moving to a fresh address with a new key.
If you want to bring your holdings together in one place anyway, check the terms of the licensed venues first: which provider lists SHIB at all, and what a transfer there costs, decides whether the move is worth it for a holding worth tens of euros.
For investors in Germany, theft has an uncomfortable tax side. Gains from selling crypto assets fall under private disposals as set out in section 23 of the German Income Tax Act. After a holding period of one year the gain stays tax-free; below that it counts towards the rest of your income, and since 2024 the exemption limit has stood at 1,000 euros per calendar year. If that limit is exceeded, the entire gain becomes taxable, not just the part above it.
Theft, however, is not a sale. The disposal is missing, and on the line taken in the German finance ministry's guidance on crypto assets no loss arises that could be offset against gains from other private disposals. The holding is gone and the tax burden on earlier gains remains. That is why the half hour spent revoking old approvals pays off more reliably than any attempt at rescue afterwards.
The duty to keep records applies regardless. Anyone holding SHIB across several addresses and chains should keep acquisition dates cleanly, otherwise the one-year period cannot be proven in case of doubt. Tools that handle this automatically read the purchases straight from the addresses. Classifying an individual case remains a matter for your tax adviser, and this text does not replace one.
From this newsroom's point of view, the timing of this trick in this particular week is no coincidence. The evidence is in plain sight: SHIB stands 93.8 percent below its 2021 high, down 49.9 percent over the year, and daily turnover of $63.1 million is thin for an asset with a $3.14 billion market capitalisation. In such an environment, a promise of free tokens meets holders who have been waiting months for a countermove. Mazrael made the same observation in his post when he spoke of a flood of such accounts.
Against that, the trick has not changed technically. Approvals as a route of attack have existed for years, and every wallet now displays warnings. Anyone going through the approval list for their addresses once a quarter removes the basis for this attack, no matter how convincing the next page looks. No statement about the further course of the price follows from this, and this section is not a recommendation to buy or sell.
(As of October 12, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
IMF chief Kristalina Georgieva says AI investment is holding global growth near 3%, with Asian chipmakers winning big, while warning of a 'slow-moving tsunami' for young workers.
The bank's Dubai branch now offers custody for Ripple's RLUSD stablecoin and plans a link with LMAX that lets institutions post it as trading collateral while the coins stay put.
Meta's new flagship beats GPT-6 Astra and Claude on memory, legal and marathon coding tests, but independent rankings still put Anthropic's Opus 5.5 on top.
REA, a free toolkit that hands AI agents the keys to reverse engineering, crossed 100,000 GitHub stars in a week as AI-decompiled game mods, console ports, and Adobe clones flood the internet.
The Nasdaq-listed company now holds about 2.59 million SOL, but its weekly purchases keep shrinking.
Multiple assets are above their long-term moving averages, which might not save the market's outlook for the future.
The U.K. High Court has delivered a pivotal ruling in the battle over 61,000 seized Bitcoin, potentially denying thousands of Chinese fraud victims access to the cryptocurrency's massive price gains.
Bitcoin is facing mounting selling pressure as the U.S. Dollar Index (DXY) surges to an 18-month high, with bearish derivatives activity and a wave of long liquidations adding to concerns about the cryptocurrency's near-term outlook.
Peter Schiff has declared Michael Saylor’s Bitcoin strategy a failure in 2026.
Cardano logs a historic 1,780,896 transactions in Epoch 660, marking its highest network usage since late 2021.
CME Group (CME) stock fell 1.18% to $277.24 on Monday, October 12, losing $3.32 despite progress on its cloud migration. The company completed initial performance testing for CME Globex on Google Cloud, achieving faster trading response times. Meanwhile, CME Group confirmed an accelerated construction schedule for its Aurora Cloud Region, supporting its planned market migration.
CME Group Inc., CME
CME Group shares reversed earlier gains during Monday’s session as selling pressure increased throughout trading. The decline came as the company announced progress toward modernizing its global derivatives trading infrastructure. However, the announcement did not prevent the stock from ending the session lower.
CME Group completed initial performance tests for CME Globex, its electronic trading platform, using Google Cloud infrastructure. The tests achieved faster round-trip times, which measure how quickly trading systems process and return messages. Consequently, the results marked progress toward delivering ultra-low-latency trading through cloud technology.
The company also reported faster construction progress at its Aurora Cloud Region and co-location facility. Google Cloud will provide specialized infrastructure designed to support the trading requirements of CME Group’s markets. CME Group expects the Aurora facility to become available for client testing by mid-2027.
CME Group Chairman and CEO Terry Duffy said the migration program continues to advance faster than initially expected. He attributed the accelerated schedule to construction progress and strong results from initial performance testing. Furthermore, the company expects its cloud infrastructure to support ultra-low-latency trading services ahead of its original schedule.
CME Group plans to move its livestock markets to the Aurora Cloud Region during late 2027. Previously, the company considered an interim relocation to Dallas before completing the transition to Aurora. However, faster construction progress eliminated that requirement and allowed CME Group to simplify its migration plans.
CME Group’s Dallas disaster recovery facility has supported client testing since June 2026. The facility remains part of the company’s broader infrastructure modernization program. Additionally, CME Group expects clients to access Google Cloud-managed services and co-location capabilities in Aurora during late 2027.
CME Group operates one of the world’s largest derivatives marketplaces, serving participants across several major asset classes. Its exchanges provide futures and options tied to interest rates, equities, currencies, cryptocurrencies, energy, agricultural commodities, and metals. The company also operates BrokerTec, EBS, and CME Clearing across its trading and clearing businesses.
CME Globex supports electronic futures and options trading, making reliable infrastructure essential to CME Group’s daily operations. The Google Cloud migration aims to combine specialized computing systems with the speed requirements of electronic markets. The transition will provide clients with managed cloud services and physical co-location options.
The latest milestones strengthen CME Group’s planned transition while establishing clearer testing and deployment targets for 2027. However, Monday’s stock decline contrasted with the operational progress outlined in the company’s announcement. CME Group will continue preparing its Aurora infrastructure as it advances the next stages of cloud migration.
The post CME Group (CME) Stock Falls as Google Cloud Migration Accelerates appeared first on Blockonomi.
Hims & Hers Health (HIMS) stock declined 0.91% to $30.64 on Monday, despite the company’s expansion into Australia’s women’s healthcare market. Shares lost $0.28 during regular trading before slipping another 0.13% to $30.60 in after-hours trading. Meanwhile, the company launched its Hers platform nationwide, expanding access to personalized healthcare services for Australian women.
Hims & Hers Health, Inc., HIMS
Hims & Hers announced the official launch of Hers in Australia on October 12, expanding its international healthcare operations. The platform offers treatments covering weight management, hair regrowth, menopause, and perimenopause through digital consultations. Women can access healthcare professionals, medication support, and a dedicated care team available around the clock.
The expansion follows the company’s acquisition of Eucalyptus earlier this year, strengthening its position in Australia’s digital healthcare sector. Under the transition, Eucalyptus’ existing women’s healthcare platform, Juniper, will operate under the Hers brand. Meanwhile, the earlier rebranding of Pilot to Hims in August completed the company’s broader Australian platform rollout.
The company introduced specialized weight management programs that combine medical treatment with nutritional guidance and fitness support. Furthermore, Hers provides hair regrowth treatments and services addressing hormonal health concerns throughout different life stages. The platform also allows patients to request female practitioners and access a community exceeding 27,000 women.
Hers plans to address healthcare access challenges affecting women across Australia’s metropolitan and regional communities. According to company research, 52% of surveyed Australian women reported experiencing weight-related judgment from healthcare professionals. Among those respondents, 72% subsequently avoided seeking clinical assistance, highlighting barriers to routine treatment.
The company also identified healthcare costs and geographic limitations as obstacles to accessing specialized medical services. Nearly 30% of Australians live in regional and rural communities, where specialist appointments can require extended waiting periods. Hers aims to connect patients with medical professionals through remote consultations and continuous digital support.
Rebecca Holdt, the company’s international product vice president, emphasized digital healthcare’s role in improving nationwide access to treatment. Meanwhile, Australian Chief Medical Officer Matt Vickers highlighted the need for evidence-based services addressing women’s broader healthcare requirements. Existing Juniper patients will retain their Australian practitioners while gaining access to additional treatment options.
The Australian launch extends Hims & Hers’ existing international presence across North America and Europe. Furthermore, the company enters a growing telehealth market with increasing demand for accessible digital healthcare services. Industry projections cited by the company estimate Australia’s telehealth market could reach $2.56 billion by 2034.
Gus Wood will oversee Australian operations as general manager, supported by Chief Medical Officer Matt Vickers. Hers Chief Medical Officer Jessica Shepherd will provide clinical oversight to maintain the company’s international healthcare standards. The leadership structure combines local medical expertise with the company’s wider technology and treatment infrastructure.
Despite the expansion, Hims & Hers Health stock ended Monday’s session lower, extending its losses into after-hours trading. The company’s latest rollout broadens its healthcare offerings beyond existing services and establishes a unified brand across Australia. However, Monday’s stock decline contrasted with the company’s continued expansion into international digital healthcare markets.
The post Hims & Hers Health (HIMS) Stock: Slides Despite Major Australia Expansion appeared first on Blockonomi.
IREN Limited stock closed Monday at $34.80, down 1.11%, before recovering 0.86% to $35.10 in after-hours trading. The movement followed reports linking the company to an undisclosed multiyear cloud agreement with Anthropic. The development renewed market interest in IREN’s expanding artificial intelligence infrastructure business.
IREN Limited, IREN
The stock had gained approximately 8% in overnight trading on Sunday following speculation about the partnership. Monday’s regular session erased those early gains as shares finished below the previous closing level. The reported agreement could strengthen IREN’s position in the growing market for high-performance computing infrastructure.
The Information reported that Anthropic maintains an unannounced multiyear cloud agreement with IREN, citing people familiar with the arrangement. The report examined Anthropic’s efforts to expand data center capacity and meet growing computing requirements. Neither company has publicly confirmed the partnership or disclosed its financial terms.
The report also connected Anthropic’s infrastructure expansion to several companies with backgrounds in Bitcoin mining. These included Hut 8, TeraWulf, IREN, and Nscale, which provide computing infrastructure and access to large power resources. Such providers can help technology companies expand capacity without developing every facility independently.
Anthropic’s reported relationship with IREN would support the company’s transition from cryptocurrency mining toward artificial intelligence cloud services. Growing demand for advanced computing infrastructure continues to create opportunities for specialized data center operators. A confirmed agreement would add another prominent customer to IREN’s expanding business network.
In August, IREN disclosed several multiyear cloud agreements during its fiscal fourth-quarter earnings call. CEO Daniel Roberts identified customers including Cohere, Prometheus, Perplexity, Figure AI, Fal AI, and Higgsfield AI. He also mentioned a separate agreement with a leading frontier artificial intelligence laboratory without naming the organization.
The earlier disclosure created a connection between IREN’s existing contracts and the latest Anthropic report. IREN has not established whether Anthropic represents that previously unnamed customer or a separate agreement. The reported partnership therefore remains an unconfirmed development rather than a newly announced company contract.
IREN has expanded its infrastructure business through agreements with technology companies seeking additional computing capacity. Its broader customer network and existing facilities support its move into specialized cloud services. The company also faces competition from CoreWeave, Nebius, and other operators expanding their artificial intelligence infrastructure businesses.
IREN has faced criticism concerning the operational quality of its older data centers in British Columbia. Research firm SemiAnalysis recently highlighted reported customer complaints involving facilities in Prince George and Mackenzie. These concerns introduced additional scrutiny of IREN’s infrastructure as the company expands its artificial intelligence operations.
Despite those concerns, retail sentiment strengthened following reports connecting IREN with Anthropic. The stock reached the top of Stocktwits’ trending list as traders discussed the possible partnership. Some market participants anticipated further gains, while others questioned the lack of official confirmation.
IREN shares have also struggled against several competing cloud infrastructure stocks during 2026. The company entered Monday’s session with a year-to-date decline of approximately 7%, according to the earlier market report. The Anthropic speculation lifted trading interest, Monday’s lower close showed that the initial rally had faded.
The post IREN Limited (IREN) Stock: Eyes Big Gains on Anthropic Deal Buzz appeared first on Blockonomi.
Navitas Semiconductor Corporation (NVTS) stock fell 7.33% to $9.99 at Monday’s close before rebounding 6.21% to $10.61 after hours. The recovery followed developments surrounding its latest AI data center power technology, developed alongside NVIDIA. The company introduced a 24 kW power delivery board designed to improve efficiency across next-generation computing infrastructure.
Navitas Semiconductor Corp, NVTS
Navitas Semiconductor announced its latest 800 V-to-12 V power delivery board for NVIDIA’s MGX ecosystem. The technology uses GaNFast components to support advanced power systems across high-performance AI data centers. Furthermore, the development expands Navitas’ existing portfolio of power solutions for demanding computing applications.
The new board delivers 24 kW while targeting 97% efficiency under full operating loads. It also achieves power density exceeding 2,000 watts per cubic inch, supporting increasingly demanding AI hardware. Consequently, the system delivers substantial power within a compact design suited for modern computing infrastructure.
Navitas developed the board using NVIDIA’s latest 400 mm by 60 mm form factor. This configuration allows manufacturers to position power components closer to processors and reduce transmission losses. Additionally, the design supports higher computing demands without requiring larger power delivery systems.
The technology converts 800 VDC directly into 12 VDC through a single-stage power architecture. This approach removes an intermediate conversion stage commonly found in existing power delivery systems. As a result, operators can reduce energy losses, hardware requirements, and overall system complexity.
The board incorporates 16 NV6034 GaNFast transistors within its primary power conversion architecture. These components use a stacked half-bridge configuration and Navitas’ latest dual-cooled semiconductor packaging. Together, these features support efficient power delivery and improve thermal management during demanding computing operations.
Navitas previously introduced a 20 kW power board that converts 800 VDC into 6 VDC. The latest development extends that earlier collaboration with NVIDIA and introduces another option for system designers. Meanwhile, both solutions support the industry’s transition toward higher-voltage power infrastructure for AI computing.
The latest announcement supports Navitas’ broader transformation strategy, which focuses on advanced power semiconductor technologies. Through this initiative, the company aims to replace traditional silicon components with gallium nitride and silicon carbide alternatives. These materials support higher efficiency and greater power density across demanding electrical applications.
Navitas President and CEO Chris Allexandre identified efficient power delivery as an important requirement for expanding AI infrastructure. He also emphasized the company’s work with NVIDIA to simplify power conversion between electrical systems and GPUs. The new technology addresses growing energy requirements across increasingly powerful computing platforms.
Navitas plans to display the new board at the NVIDIA AI Factory MGX Wall during OCP Global Summit 2026. The presentation will highlight its latest power delivery capabilities within NVIDIA’s expanding computing ecosystem. Meanwhile, the stock’s after-hours rebound followed a difficult regular session as the company advanced its AI infrastructure strategy.
The post Navitas Semiconductor Corporation (NVTS) Stock: Rebounds as NVIDIA Unveils New AI Power Tech appeared first on Blockonomi.
Intel Corporation (INTC) stock fell 3.20% to $101.35 at Monday’s close before recovering 0.22% to $101.57 in after-hours trading. The decline came as Intel revealed new GPU specifications and Xeon server designs at the Open Compute Project Global Summit. The announcements highlighted Intel’s efforts to expand its data center business through efficient artificial intelligence infrastructure.
Intel Corp., INTC
Intel introduced further technical details for its GPU D-series, previously known as Crescent Island, during the OCP Global Summit. The new processor uses Intel’s Xe3P architecture and targets memory-intensive AI inference across enterprise data centers. Meanwhile, the company aims to improve processing efficiency while reducing infrastructure costs and power consumption.
The GPU architecture supports up to 480GB of LPDDR5X memory and delivers 1.5TB/s of memory bandwidth. Additionally, the design provides 1.3 petaflops of FP8 computing performance within a 350-watt power limit. These specifications support large language models, long-context processing, and multiple inference workloads running simultaneously.
Intel said a single GPU can support models containing up to 300 billion parameters. Furthermore, an eight-GPU configuration can handle models exceeding one trillion parameters. Customer sampling has started, while Intel continues collaborating with hardware manufacturers and software providers on deployment solutions.
Intel also presented new Xeon-based rack designs that support growing computing requirements across enterprise AI systems. The designs use open infrastructure standards, including Open Rack v3, to simplify integration into existing data centers. Consequently, organizations can expand processing capacity without completely replacing their current infrastructure.
Select Xeon 6+ configurations deliver more than 46,000 processor cores per rack. These systems operate at 133.45 kilowatts of total rack power while supporting high-density computing operations. Intel designed the configurations to handle data processing, workload coordination, and communication between specialized accelerators.
Agentic AI systems require processors to manage several connected tasks rather than complete isolated requests. Therefore, Intel positions Xeon processors alongside GPUs to handle workflow management and data movement. This approach supports concurrent AI applications while improving resource allocation across enterprise computing environments.
Intel outlined additional software improvements that support applications across its CPU and GPU product lines. The company continues integrating established AI frameworks, inference tools, and open development environments into its platforms. These efforts aim to simplify application deployment and reduce software compatibility barriers.
Intel also highlighted its contributions to the Open Compute Project, including standardized accelerator card designs. Since 2011, the company has submitted 116 contributions supporting open data center infrastructure. Its latest work focuses on improving hardware compatibility and giving system manufacturers greater design flexibility.
The announcements extend Intel’s strategy of combining general-purpose processors, specialized accelerators, and interoperable software. Meanwhile, the company plans further technical presentations during the OCP Summit on October 13. Its expanding hardware portfolio targets enterprise demand for scalable AI processing and more efficient data center operations.
The post Intel Corporation (INTC) Stock: Slides as New AI GPU Details Emerge appeared first on Blockonomi.
HTTP error 429 on https://cryptopotato.com/feed/
Failed to fetch feed.