Cryptopolitan

Danaher Corporation has announced that it will build its first AI-powered autonomous laboratory meant to design, build, and test custom antibodies with minimal human intervention and with plans to reach full operating capacity in early 2027.
Danaher supplies different instruments and reagents for drug developers and academic researchers, and the new lab is aimed at the expectation that automating the most time-consuming areas of early-stage discovery will give the company an edge commercially.
The lab will be based at Abcam, one of Danaher’s operating companies, and combine AI, robotics, and tools from across the group in a single workflow. The system will use AI to design new affinity reagents, including antibodies, then have robots build and test the candidates. The results will feed back into the model, allowing each new round of designs to improve on the last, according to Danaher’s announcement. The company has termed this as a continuous “design-make-test-learn” loop.
Five Danaher units will supply the underlying technology, including the Beckman Coulter Life Sciences, Cytiva, Genedata, Integrated DNA Technologies and Molecular Devices. The robotic automation and device orchestration was accessed via work with Automata, a firm Danaher invested in back in January 2026, according to Quartz.
Eight times faster with ten times more output
According to the company’s press release, the lab is expected to speed up affinity reagent discovery by up to eight times and eventually increase annual reagent output by almost ten times, moving from tens to hundreds as it scales over time.
JC Gutierrez-Ramos, Danaher’s senior vice president and chief science officer, said the lab represents a new way of doing scientific research, combining AI and automation with human expertise in a continuous cycle. He added that scientists would still make the key decisions, while the autonomous platform would handle the repetitive experimental work.
Danaher CEO Julie Sawyer Montgomery described the project as an early step and not a finished system. She said it could help usher in a future where AI, automation, and intelligent systems contribute hugely to speed up the path from scientific discovery to real-world impact.
Danaher pitches one lab with a larger plan
The Abcam facility is expected to be the first step in a continuous effort to build what the company calls a fleet of smart instruments. These devices are designed to be easier to control through software, generate data that is ready for AI use, and run “expert-level” agentic systems.
The instruments in use could eventually be connected across more autonomous labs, allowing the same self-improving cycle to run across different research programs, per the announcement.
Investors reacted slightly positively, with shares rising by about 2% in premarket trading on Wednesday, according to Yahoo Finance data. Before the announcement, Barclays analyst Luke Sergott had raised his price target for Danaher to $255 from $230, while keeping an Overweight rating on the stock.
Danaher shares were trading at around $219 as of the time of writing.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Google made its SynthID Detector available in English to everyone in the world on October 7.
The free tool looks through pictures, videos, and audio for a watermark that you can’t see. Google has put this watermark on more than 180 billion pictures, videos, and 240,000 years of audio.
There are already checkers built into Chrome, Search, and the Gemini app that do more than a million checks every day. So now the detector has its own site called synthid.com next to them.
Gemini’s built-in AI check had been used 50 million times before May
Up until now, only researchers, journalists, and people who work in the media had access to an early version that Google gave out last year. Now, anyone can upload a file and find out if it was made by a participating model.
Google added SynthID checks to Lens, AI Mode in Search, and Chrome at its I/O event in May 2026. The check had been run 50 million times in the Gemini app by that point.
In the notice, Pushmeet Kohli said the goal is to give people “more context about the media you see online.” He is VP of science and strategic initiatives at Google DeepMind.
In 2023, Google DeepMind started with images. Later, they added text, video, and audio to SynthID. In 2024, they made the text version open source.
The mark stays inside the content, in the pixels of an image or video frame, or the waveform of an audio file. This means that it never interacts with the file’s metadata. Veo, Lyria, Nano Banana, Gemini, Flow, ProducerAI, and Vids are all Google’s own generators that add it to their output.
Image types supported by the site include JPG, PNG, WEBP, GIF, and HEIC. Video types supported include MP4, MOV, and WEBM. Audio types supported include MP3, WAV, and FLAC.
The detector reads OpenAI, Nvidia and Kakao marks
Google lists OpenAI, Nvidia, and Kakao as partners and says Apple will be joining them soon. ElevenLabs also supports SynthID.
The SynthID Detector homepage at synthid.com.
OpenAI also has its own site that checks for content. Microsoft and Meta have different rules for watermarking and verifying content, and their tools often miss content that was made by their own models.
If the answer is “yes,” the file is linked to Google or one of its partners. The detector comes back empty when it comes across output from AI systems that are not in that group.
A negative result only means that there was no SynthID signal. It doesn’t prove that someone made the file.
Google adds the C2PA Content Credentials on top of the hidden mark. This is an industry standard that keeps track of where a file came from and how it was changed.
Apple is going to add SynthID to its image tools in iOS 27. Cryptopolitan reported in August that Apple is also testing its own method in the iOS 27 beta. This is done with a Reference Image feature that connects a picture to the camera sensor that took it.
The smartest crypto minds already read our newsletter. Want in? Join them.

Elon Musk has said Tesla and SpaceX will build and operate Terafab, the semiconductor plant he is developing in Texas, putting to rest speculation that Taiwan Semiconductor Manufacturing Co. (TSMC) could take over the project and its operations on any scale.
Musk posted the comment on X in reply to a video, days after telling followers he was in talks with TSMC about the project. He narrowed those talks to only the possibility of renting out floor space.
@herbertong @thejefflutz No, we will build and run the fab. Let there be ZERO doubt about that.
Maybe TSMC subleases part of the Terafab if they want, but nothing more than that.
— Elon Musk (@elonmusk) October 7, 2026
Intel to supply underlying technology
Intel will continue to work with Musk’s companies on the technology that will support and power Terafab, aimed at developing the systems that will be installed within the plant. This relationship has remained in place regardless of the increase in speculation surrounding a potential TSMC partnership.
Intel CEO Lip-Bu Tan told Bloomberg that the chipmaker will continue working with Elon Musk on the proposed Terafab semiconductor plant. Intel joined the project as a partner in April 2026.
The project and facility in general are expected to supply advanced chips across Musk’s businesses, from Tesla’s vehicles and robotics operations to SpaceXAI, a merger between Musk’s space business and xAI, the Grok chatbot developer.
Terafab is expected to be split between two factories, with one focused on chips for Tesla vehicles and humanoid robots and the other supplying hardware for AI data centers in space, with the project ultimately targeting an annual computing capacity of about 1 terawatt.
A crowded Texas chip map
TSMC is not stepping back from Texas totally, or completely uninterested in expanding within the state. The firm, whose customers include Nvidia Corp. and Apple Inc., is said to still be considering a new chipmaking campus in the state as it expands across the US, according to Bloomberg.
If the campus moves forward, it would add to an expanding semiconductor hub in Texas.
Samsung Electronics already has a plant under development in the state, with Terafab already promising extra additional chipmaking capacity. Reuters had also previously reported that SpaceX and Tesla committed $16.8 billion in initial spending on the Terafab project.
The smartest crypto minds already read our newsletter. Want in? Join them.

The National Bank of Kazakhstan (NBK) and the Alatau City Authority have signed a memorandum of understanding (MoU) with USDT issuer, Tether, to explore frameworks for a tenge stablecoin and real-world assets tokenization.
The deal announced by the stablecoin issuer on Wednesday, October 7, adds a new name to Kazakhstan’s push to become Central Asia’s leading crypto hub following Binance’s agreement with NBK.
Tether is exploring a tenge stablecoin and tokenization framework
According to Tether’s announcement, the stablecoin issuer is looking to develop two tracks in the agreement it penned with NBK and ACA.
First, it will examine international models to develop a concept and a pilot proposal for a stablecoin pegged to Kazakhstan’s national currency, the tenge.
On the second track, they are working with domestic authorities to build a tokenization framework for real-world assets, potentially using Tether’s Hadron tokenization platform.
Tether CEO Paolo Ardoino said, “Kazakhstan is taking a deliberate, forward-looking approach by exploring asset tokenization and a strategic reserve framework.”
Financial stability, transparency, and investor protection are the central bank’s top priorities in this arrangement, according to a deputy governor at the National Bank, Binur Zhalenov.
Both sides also committed to a run of workshops and training sessions for central bank and government staff, covering stablecoin issuance, reserve management, and tokenization.
Why firms are moving to Alatau, Kazakhstan
Alatau, the city run by the Alatau City Authority (ACA), is Kazakhstan’s “CryptoCity” ran under a special legal regime signed into law on May 8, 2026, by President Kassym-Jomart Tokayev. Deputy Prime Minister Kanat Bozumbayev also called Alatau a “charter city” during a May 21 briefing.
Kazakh outlet Qazinform reports that the special legal regime framework, partly modeled after Dubai and Singapore, took effect in stages starting July 1.
Disputes can be resolved through the Astana International Financial Centre (AIFC) or international arbitration, and local rules can draw on the law of England and Wales.
How Alatau is set up.
Notably, Bozumbayev said the successes from CryptoCity could be deployed on a larger scale throughout Kazakhstan.
Tether is following Binance to Alatau, after the exchange signed a September 7, 2026, memorandum with the National Bank of Kazakhstan to build a regional settlement hub in the country, as Cryptopolitan reported at the time.
Binance had a local presence even before that September MoU, as its Binance Pay service went live across 5,000 point-of-sale terminals through Alatau City Bank. Co-founder Changpeng Zhao (CZ) announced the milestone on July 8, 2026. That system converts digital assets, mostly USDT, into tenge instantly for merchants.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Genius Group, the Singapore-based education company that once held 440 bitcoin, has started buying again.
The company will be starting its reserves from scratch after it sold all 440 coins it held to clear debt.
Is Genius Group holding Bitcoin?
Genius Group (NYSE American: GNS) said on October 6 that it acquired 10 bitcoin for roughly $854,000. This is the company’s first Bitcoin purchase since it emptied its treasury to pay off its debts. It is also the first purchase since a U.S. appeals court cleared away the legal order that had frozen the firm’s fundraising for most of two years.
The 10 coins were bought between October 2 and October 5 at an average of $85,364 each. Chief Executive Roger James Hamilton called the purchase “modest” but said it is to signal to shareholders and to the market that the company has resumed its Bitcoin accumulation strategy.
Genius Group describes itself as an AI-powered education group serving 6 million users across more than 100 countries. It has run a “Bitcoin-first” policy since late 2024, when its board voted to hold 90% or more of reserves in bitcoin and endorsed the corporate-treasury approach popularized by Michael Saylor’s Strategy.
The company’s treasury reached 440 bitcoin by February 2025, spending $42 million at an average of $95,519 per coin. Days later a temporary restraining order and preliminary injunction, granted in a dispute with LZG International, barred the company from issuing shares, raising capital, or buying bitcoin.
During that freeze, the company was forced to offload most of its holdings. It briefly resumed buying Bitcoin in May 2025 after the order was suspended and rebuilt its holdings to 85.5 coins, but soon enough on April 1, 2026, Genius Group sold what remained to repay $8.5 million in debt in full, leaving it with no crypto reserves.
Genius Group is restarting its Bitcoin treasury from zero. Source: BitcoinTreasuries.net.
The company said at the time it would return to the market “when it believes market conditions are more favourable.”
Why can Genius Group rebuild its treasury now?
On August 31, 2026, the Second Circuit Court of Appeals overturned a preliminary injunction that a New York federal court had issued against Genius Group in March 2025. The company says that order was based on false and inaccurate statements and added that this ruling ends an 18-month legal battle, letting it carry out its treasury plans without legal limits.
Separately, Genius Group is pursuing a racketeering (RICO) lawsuit in Florida against LZG officers and other defendants, seeking more than $750 million in triple damages.
The 10 bitcoin purchase is the opening move in a $1.2 billion dual treasury plan the board approved and announced on August 27. The plan targets $827 million in bitcoin and $800 million in AI assets, with a goal of $2 billion in total assets by fiscal 2031.
On the AI side, the company reports look-through stakes in OpenAI, Anthropic, Databricks, and SpaceX.
Genius Group says it will pay for the purchases using cash from its operations, selling new shares gradually at market prices (an at-the-market share offering), and a planned perpetual preferred security. It has no current plans to take on debt or to use its bitcoin and AI holdings as collateral.
This approach is similar to Strategy’s. The company launched its preferred stock (called STRK) in January 2025 and has raised over $16 billion with it.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Microsoft will launch its Surface Laptop Ultra line of laptops at a San Francisco event on Wednesday, a new line of premium Windows PCs using Nvidia’s RTX Spark chip and built to support AI agents directly on the device instead of operating them on the cloud.
Microsoft CEO Satya Nadella and Nvidia chief Jensen Huang are set to take the stage together to unveil the machine, Reuters reported. The laptop is said to be the product of years of collaboration between the two companies.
Pushing AI work off Azure and onto local devices
Today a lot of Microsoft’s AI runs within its Azure data centers, an expensive proposition when viewed at scale. By moving tasks such as code writing and complex business work onto Windows laptops that customers purchase, Microsoft shifts part of that compute bill to its users, according to Reuters.
The company still holds a commanding share of the PC market, which makes it a natural area for the experiment. The partnership also offers a way into a Windows PC market that has been dominated by Intel and AMD for Nvidia.
The RTX Spark chip was revealed in June, and is an Arm-based system-on-chip developed along with MediaTek that brings together the CPU, graphics and unified memory all into one package.
What Microsoft has confirmed, and what leaked
Microsoft’s product page highlights the hardware behind the new machine. It has a 15-inch mini-LED PixelSense Ultra touchscreen with a 3:2 aspect ratio and peak HDR brightness of 2,000 nits, while the chassis is less than 18mm thick and weighs less than 4.5 pounds.
Ports include USB-C, USB-A, HDMI and a full-size SD card reader. The touchpad is about 30% larger than the previous model, and the SSD can be replaced by the user. The thermal system also has up to 2.5 times the heat capacity of the 15-inch Surface Laptop 7th Edition. Microsoft also stated that the chip can deliver up to one petaflop of AI compute and support up to 128GB of unified memory.
Leaks suggest the laptop will come in two configurations. The base version is expected to pair an 18-core CPU with 5,120 CUDA cores and 24GB of LPDDR5X memory, while the higher-end model would step up to a 20-core CPU and 6,144 CUDA cores. The two chips will also differ in memory capacity, with the smaller version reportedly capped at 32GB and the larger one supporting up to 128GB.
The launch’s price problem
Pricing is expected to start above $2,000 for the new Surface Ultra device, with the highest-end configurations potentially reaching $7,000 to $8,000 when equipped with 128GB of memory. Microsoft has not yet disclosed official prices for the devices. Anshel Sag, an analyst at Moor Insights & Strategy, told Reuters that rising memory costs are causing a massive headache for both Microsoft and Apple. He explained that the software is ready, but the hardware needed to run AI locally remains too expensive. This means local AI is still largely limited to people who can afford high-end hardware.
A memory-chip shortage has driven up component costs across the industry, and AI machines have been hit the hardest, compared to two years ago, when Microsoft first began pushing the idea of running AI workloads directly on PCs. Laptops with those capabilities generally cost less than $2,000 at the time, but the prices for these have since risen sharply.
Keeping agents contained, amid other issues
Microsoft and Nvidia also need to convince buyers that AI agents, which can work autonomously for extended periods, can operate safely without escaping the confines of the PC.
The hardware powering the new Surface laptops is not Microsoft’s alone. At IFA last month, Nvidia said the first RTX Spark laptops and desktops would land in October, with Dell, HP, Asus, MSI, Lenovo, and Acer all backing the platform. Windows and Devices chief Pavan Davuluri is expected to join Nadella and Huang at the event, which is set to involve more extensive updates on Windows and the Surface line.
Ahead of the event, Microsoft shares were down 0.47% at $526.83, while Nvidia fell 1.05% to $236.72.
The smartest crypto minds already read our newsletter. Want in? Join them.

The Cardano Foundation is getting ready for a future where issuers of regulated tokens can freeze or seize stablecoins, funds and bonds issued on the blockchain, as confirmed by the compliance toolkit.
Per Cardano, the toolkit that went live on mainnet on Wednesday will allow issuers to screen recipients and also block sanctioned wallets on the network. The entire stablecoin market, which is the sector that uses the function most often, is relatively small on Cardano, at roughly $67 million.
According to the Swiss-based nonprofit that stewards the network’s development, the CIP-0113 Cardano improvement proposal it announced at the TOKEN2049 conference went live without requiring a hard fork.
Can tokens be frozen on Cardano now?
The CIP-0113 that went live on Cardano is a compliance tool being pitched as a compliance logic that follows tokens everywhere, and every time, checking that tokens continue to meet the rules by issuers.
Issuers can set parameters, including identity and AML checks, screening for sanctions, transfer restrictions, and freeze-and-seize controls from pre-built rule sets, called modules. They can also write their own rules and can swap them out as regulations shift without touching the underlying standard.
How it works in real life is that, whether it is a transfer, mint or burn, the Cardano ledger will first check that the transaction meets the issuer’s rules before approving it. Transactions that don’t meet the issuer’s standards are rejected.
In cases where tokens are restricted, those bars are limited to just one app; they follow them across wallets and services.
Cardano Foundation’s chief executive, Frederik Gregaard, said: “The rules have to travel with the asset and be enforced every time it moves.”
The global powers that CIP-0113 hands to token issuers also come with a headache. For example, if the issuers’ rules allow it, it can claw back tokens from a user’s wallet without their consent. That feature is specifically why the technical specification warns lending services to double-check that the issuers of the tokens they accept as collateral don’t have those powers.
Notably, the “Freeze & Seize” feature is an optional regulatory compliance feature according to the GitHub repository for the Aiken implementation.
Cardano’s stablecoin market is relatively tiny compared to the competition
Relative to Ethereum, Solana, and XRP Ledger, which already have their own versions of transfer controls, Cardano stablecoin market is relatively small, counted at roughly $67.5 million per DeFiLlama data.
Top stablecoins on Cardano. Source: DeFiLlama.
Cardano arrived at the Coinbase x402 payment standard relatively late, only adding ADA support in September, as Cryptopolitan reported.
CIP-0113 is Cardano’s answer to the Bank for International Settlements (BIS) and International Monetary Fund (IMF)’s call to add programmability, the ability to hard-code compliance into an asset, as a priority requirement for the growing tokenized finance market.
The Swiss Capital Markets and Technology Association gave the thumbs up on CIP-113 Programmable Asset Tokens, saying it recognizes it as a smart contract equivalent to its CMTAT framework, usable for certifying compliance of ledger-based equity securities under Swiss standards.
The work is older than this week’s headline. The implementation builds on an earlier reference design, CIP-143, originally developed by Phil DiSarro and the IOG team and since migrated to the Aiken language.
Back in January 2025, Cardano founder Charles Hoskinson had taken an interest in the proposal, framing freeze-and-seize as a tool for non-adversarial uses too, such as asset recalls, identity updates and dividend payments. The Foundation says a dedicated securities module for regulated instruments is still in development.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

Advanced Micro Devices (AMD) is still considering a foundry partnership with Samsung Electronics, CEO Lisa Su said in Seoul on Wednesday, as the chipmaker works to secure the memory supply needed for its next generation of AI accelerators.
Su met with Jun Young-hyun, Samsung’s vice chairman and head of its Device Solutions division, at the company’s Seocho office in Seoul’s Gangnam district on Wednesday. It was their first meeting in about seven months, following their previous meeting in March. AMD senior vice presidents Jack Huynh and Joseph Macri were also in attendance.
Su kept the agenda on ground non-specific when reporters asked her for details, saying the conversation would focus on AMD’s wider partnership with Samsung. She also noted that AMD has multiple partners within the Korean electronics giant. When asked if she would also meet Samsung Executive Chairman Jay Y. Lee, Su declined to elaborate, replying, “Too much information,” the Seoul Economic Daily reported. Earlier that day, she met SK hynix President Kwak Noh-jung, marking their first meeting in Korea, according to Businesskorea.
HBM4 is why AMD keeps coming back to Korea
The pull for AMD towards Korea is memory, and specifically high-bandwidth memory. AMD’s newest AI accelerator, the Instinct MI455X, carries 432GB of sixth-generation HBM (HBM4), a 50% increase on the 288GB of HBM3E in the earlier MI355X, CHOSUNBIZ reported.
Su said memory supply remains “very tight,” with the chipmaker working with customers to use memory more efficiently while pushing suppliers to increase output as quickly as possible.
AMD had placed Samsung as its preferred HBM4 supplier for the MI455X accelerator back in March, and Su told reporters at an “AMD Gamer Day” event in Seoul that shipments of the MI455 and the Helios system, both built on the HBM4 memory, had already started.
Su also highlighted how much AMD depends on South Korean suppliers. She said the Korean supply chain is a major part of AMD’s operations, with Samsung and SK hynix playing a central role in the company’s data-center ambitions.
Foundry question amid TSMC reliance
AMD currently relies on Taiwan’s TSMC to manufacture its advanced CPUs and GPUs, including the MI455X, which is built using a TSMC process. Adding a second manufacturer would require AMD to repeat the qualification process, including checks on yields, performance and delivery timelines.
AMD also warned in its latest annual report that moving some production to a new foundry could delay product launches or potentially hurt yields.
The companies’ March memorandum still left room for a more extensive partnership. Along with HBM4 and next-generation DDR5 supplies, the two sides agreed to explore producing future chips through Samsung’s foundry, with Samsung also offering its advanced packaging capabilities.
The foundry question has continued to remain a sticking point though, with one senior semiconductor-industry figure telling CHOSUNBIZ that AMD “sees too much risk in entrusting Samsung Electronics with foundry,” suggesting the current talks are largely focused on closing the gap between the two sides.
Landing actual AMD chip orders could accelerate a turnaround in Samsung’s loss-making contract manufacturing business, the Seoul Economic Daily reported. Su stopped short of confirming any new foundry deal but left the door open. She said Samsung Electronics remains an important partner across several areas and that AMD continues to look for opportunities to work together. She also stressed the need for a broad supply chain that can be planned three to five years ahead.
The smartest crypto minds already read our newsletter. Want in? Join them.

Brussels is looking at a way to pull more money out of Apple, Google, and Meta while giving Washington no reason to retaliate against it.
The country is considering including the tech giants in a flat levy placed on every large company operating in the bloc.
Is the EU going to impose more taxes on US big tech companies?
The EU is considering using an existing proposal called the Corporate Resource for Europe, or CORE, as a vehicle to impose a flat levy on tech giants operating in the region without provoking retaliation from the U.S.
CORE requires companies above a certain size to hand over a fixed yearly payment of between €100,000 and €750,000. The European Commission is now considering rewriting that rule so it would apply the charge to any firm doing business in the EU with an annual revenue above €100 million (roughly $112.32 million).
The approach avoids naming the American companies it is partly meant to target. One EU official reportedly stated that some European capitals shy away from imposing a straightforward digital tax due to an unwillingness to antagonize the Americans. The official also pointed out that many more of these capitals object to CORE.
The hesitation is understandable, considering that in June, President Donald Trump threatened a 100% tariff on goods from any country that imposes a digital services tax on U.S. firms. He wrote that the penalty would hit “any and all Goods” and override trade deals already on the books.
That warning came shortly before a July 4 deadline linked to a framework agreed in May, which limits most tariffs on EU exports to 15%.
Washington previously called the EU’s attempts to impose duties on its big tech companies discriminatory, but the proposed levy will include European manufacturers, retailers and banks alongside Silicon Valley giants.
Why can’t Europe tax U.S. companies easily?
Europe’s finance ministers abandoned an EU-wide digital tax in March 2019, as skeptics warned it would draw Trump’s ire and dent competitiveness.
The think tank ECIPE noted that the Commission dropped a proposed digital levy from its budget plans in July 2025, again under U.S. trade pressure. In the meantime, France, Italy, Spain and Austria have gone ahead with their own national digital taxes.
In 2019, the USTR concluded that France’s digital services tax singled out Google, Apple, Facebook and Amazon and prepared duties of up to 100% on $2.4 billion of French products.
Washington has sharply criticized the EU’s regulatory crackdown on American tech, particularly following the European Commission’s €890 million fine against Google under the Digital Markets Act (DMA)—comprising €460 million for search self-preferencing and €430 million for Google Play Store anti-steering practices. Cryptopolitan reported that the fine itself is another challenge, with Google appealing EU orders to share search data and open Android to rivals.
The smartest crypto minds already read our newsletter. Want in? Join them.

Bank of Russia has authorized several entities to join the country’s recently regulated cryptocurrency market, making them the first legal providers of digital-asset services.
Two major state-owned banks are among the approved participants, a clear indication that Moscow intends to keep the expanding industry under strict government control.
Central bank clears Russia’s first crypto trading and custodial platforms
The monetary authority in Moscow has effectively permitted nine organizations to begin operations with cryptocurrencies under the Russian Federation’s new rules.
Four have been recognized as platforms facilitating the exchange of digital currencies, and five will serve as so-called “digital depositories,” which will provide custodial services.
The platforms have been added to their respective registries, the Central Bank of Russia (CBR) announced, making it clear these are the first entries in the two official lists.
In a press release on Tuesday, the financial regulator also highlighted:
“These participants gained market access under the transitional provisions of the Law on Digital Currencies and Digital Rights, which entered into force on September 1, 2026.”
After much anticipation and some delay, the legislation was passed by both houses of Russian parliament in July of this year and signed by President Vladimir Putin in early August.
Coming into full effect at the start of the following month, it legalized cryptocurrencies and key transactions with them like investment, storage and trading, as reported by Cryptopolitan.
From the date of their registration, the crypto platforms are obliged to follow Russia’s rules for processing and recording transactions involving digital assets and rights, the CBR stressed.
They are also required to bring all their operations into full compliance with the comprehensive regulatory framework by September 1, 2027, at the latest, the central bank further emphasized.
Russia’s biggest bank Sber to run a crypto depository service
While crypto-only firms, such as existing coin trading platforms, are free to apply for authorization from the Bank of Russia, the procedure has been significantly simplified for traditional financial market players such as banking institutions and stock exchanges.
Thus, digital depository licenses have now been granted to Sberbank, Russia’s biggest lender by assets and market share, as well as VTB, the second-largest bank in the Russian Federation.
The other approved custodians are the tokenization platform Atomyze, the blockchain developer Voltari, and a company called “Cloud Infrastructure.”
VTB has been added to the list of “organizations facilitating the exchange of digital currencies,” too, alongside three other firms that will be allowed to trade crypto.
These are the T-Invest Lab subsidiary of T-Bank, a major commercial and digital bank headquartered in Moscow, formerly Tinkoff Bank, as well as the companies Zefir and Sistema-crypto.
The first applications for registration were submitted October 5, when the CBR started accepting them, the Russian crypto news outlet Bits.Media and the Interfax news agency noted in their reports.
Crypto transfers to be made only through licensed providers
Under the “digital currency” law, all crypto transactions, including transfers to wallets hosted abroad, must be carried out through Russia-licensed intermediaries – banks, brokers, exchanges, and others.
“Digital depositories” were introduced as a new category of platforms that will be responsible for safekeeping and transferring coins and other assets, as well as providing access to their addresses.
Crypto exchanges are allowed to buy and sell digital currencies “in their own name and at their own expense outside of organized trading venues,” the Bank of Russia clarified, adding in a statement:
“The emergence of the first registry participants marks a significant step toward establishing a regulated infrastructure for the Russian digital currency market.”
Over the past couple of years, the previously conservative regulator significantly softened its stance on crypto legalization, in the light of fiat restrictions imposed in response to Moscow’s invasion of Ukraine.
Both majority-state-owned banks receiving cryptocurrency licenses now, Sber and VTB, as well as others involved in the digital-asset space like Atomyze, have already been targeted in Western sanctions.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.

Robinhood now holds Bitcoin on its own books and is among a growing list of BTC treasury companies. News that the trading platform had bought $25 million worth of BTC for its corporate balance sheet, its first such purchase, came out today when Robinhood’s senior vice president and general manager of crypto and international, Johann Kerbrat disclosed the buy in an interview on The Block’s The Starting Block. Kerbrat made the announcement on the same day he was speaking at Token2049 in Singapore.
Kebrat made it clear about why the company made the purchase. He told The Block that the reason was to line up the company’s vision with the crypto community. He was just as quick to play down the size of the purchase. With Robinhood valued at around $100 billion, he said the position is “not going to change a lot of the current trajectory of the company”.
The Bitcoin Buy Is About 0.025% of Robinhood’s Market Cap
Robinhood shares closed at $112 on Tuesday, leaving the company with a marketcap of roughly $100 billion. The new Bitcoin holding works out to about 0.025% of that. The buy is also smaller than what Strategy picked up last week alone. Michael Saylor’s firm bought 334 BTC for about $28.7 million between September 28 and October 4, taking its total holdings to 848,000 BTC.
Compared with what sits on Robinhood’s platform, the gap gets wider still. According to data from Arkham, the company holds 185,232K BTC, valued at $15.57 billion, in custody for customers. At least for now, this latest news reads more like a gesture to crypto users than a treasury strategy.
The Real Crypto Spending Is Going Into Perps and Robinhood Chain
Robinhood’s heavier commitments to crypto sit on the product side. Last week, the company said eligible U.S. users will be able to trade perpetual futures in the coming months, going long or short on BTC, ETH, SOL, XRP, DOGE, ADA, LINK and HYPE. CEO Vlad Tenev called it “a new chapter for U.S. derivatives” in a post on X.
Robinhood is bringing America its first true perps.
No expiry, with P&L settled every 15 minutes.
A new chapter for US derivatives.
— Vlad Tenev (@vladtenev) September 30, 2026
Robinhood Chain, the layer 2 built on Arbitrum’s tech stack, went live on mainnet in July. Its total value locked stood at $1.05 billion as of Tuesday, per DefiLlama. Kerbrat has pointed to the platform’s 28 million funded accounts, 27 million of them in the U.S., as distribution few L2s can match.
Tokenized stocks are growing fast enough to bump against regulation. In a separate interview with The Block last week, Kerbrat said Stock Token volume is already high enough to hit the caps set under the SEC’s innovation exemption for tokenized U.S. equities.
Bitcoin Trades Near $84,000 as the Purchase Lands
The timing comes during a rough patch for the market. Bitcoin briefly slipped below $84,000 on Oct. 6 as long liquidations across crypto climbed to $487 million, and is trading near $84K currently.
A $25 million buy will not move the price on its own. What matters more is whether Robinhood comes back for more. Repeat purchases on a set schedule, or a policy that ties buys to crypto revenue, would put the company in a different category from where it sits today. For now, the custody book and the product roadmap say a lot more about Robinhood’s crypto exposure than its balance sheet does.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Abstract is the latest L2 chain to join the long series of shutdowns in 2026. Which networks are winding down in the past year, and how will this affect users?
Abstract announced its shutdown in early October, but its decision is not unusual. In 2026, a long list of networks announced a faster or a more abrupt sunsetting. Just before Abstract, Blast announced it would wind down by the end of October.
The main reason for network shutdowns is diminished usage and liquidity. At its peak, Abstract secured over $291M in value, losing over 90% just before its shutdown. Abstract still carries over 93K weekly active addresses and 1.64M weekly transactions. The network will operate for bridging and migrations until December 15.
Abstract joined the list of shutdowns, instead of launching a token
Abstract was a tokenless platform, and decided against holding a TGE to salvage the chain’s value. In the past two years, Abstract has also been supported by Igloo, Inc., the creator of Pudgy Penguins.
Luka Netz, the founder of Pudgy Penguins, announced Abstract will wind down, and all attention will go toward the Penguins community.
‘After losing tens of millions of dollars over two years, building consumer products, assembling an all-star team, onboarding some of the biggest brands in the world, and building a community of millions, we still had not found product-market fit,’ said Netz in an X statement.
Abstract attempted to gain ground as a chain for consumer crypto, but stagnated due to limited DeFi projects and limited usage.
After the Abstract shutdown announcement, Polygon caused a conflict by posting its own statement that it will not shut down its network. Polygon’s founder, Sandeep Nailwal, stated he had no intention of subtweeting Abstract or directing enmity toward the project.
For now, Polygon has been saved by being used as a payments network, while also carrying traffic for Polymarket. However, Polygon still wound down its ZKEvm version, which did not manage to grow its traffic oc liquidity.
Which networks shut down in 2026?
Blast is one of the high-profile shutdowns in the past year. The network will operate until October 26 to allow users to move their funds. Coinbase announced it would still work with BLAST tokens until October 20.
Lisk is another chain that will shut down on October 31 and continue as an L2 Ethereum-based rollup. For years, Lisk operated as an independent L1, with significant costs and a highly involved process to select its 100 validators. An L2 network may have lower costs while preserving the Lisk brand.
Other notable shutdowns include Kadena, which faced decreasing developer activity and usage. Harmony Protocol ceased core development and active chain maintenance after the $100M Horizon Bridge hack.
Zero Network shut down due to low usage and practically no economic activity. Sophon shut down for the same reason, including its unsustainable gas economics.
Other chains like Loopring and Movement shut down their original networks and attempted a restructuring with new technologies.
The network shutdowns are not necessarily a negative sign for L2 usage. In the past months, the share of L2 economic activity recovered to over 30%.
L2 network economic activity actually increased in 2026, but niche or idle networks shut down as users concentrated into a handful of the most liquid chains. | Source: GrowThePie
Improved app revenues came from a handful of widely used chains, especially Arbitrum. At the same time, liquidity did not spread as intended through cross-chain technology, leaving some networks with virtually no activity.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Security teams that join Anthropic’s expanded Cyber Verification Program must agree to let the company retain their data so it can watch for cyber misuse.
Anthropic rebuilt the program Tuesday into three tiers of access with fewer cyber blocks at each level.
Anthropic will offer zero data retention later this fall
Enterprise Frontier Safeguards, coming later this fall, will combine the same safeguards with zero data retention. Eligible organizations will then be able to hold data in cloud infrastructure they control.
Until then, organizations with zero-data-retention access to Claude Fable 5.1 or Claude Mythos 5.1 can join without retention.
The program is available on the Claude Platform, Google Cloud’s Vertex AI, and Microsoft Foundry. On Amazon Bedrock, access is limited to customers eligible for Enterprise Frontier Safeguards.
Defense Access is the first level and was made for responding to incidents, breaking down malware, and making sure that discovered bugs are real.
Security teams, open-source maintainers, researchers who have reported bugs in the past, and critical infrastructure operators as small as a regional hospital can all apply to join the Defense Access tier. Anthropic says it will reply to applicants in a few days.
Red Team Access focuses on authorized penetration testing and is exclusively available to organizations. Anthropic indicates that the application vetting process may take several weeks.
Red Team users get cut off mid-task if they try to deploy ransomware, damage physical systems, or pen test high-risk safety systems.
The most relaxed rules belong to Specialized Access. Only a small group gets in, such as organizations cleared to test flight systems, power grids, telecom networks, or the systems banks use to move money between each other.
Anthropic checks every Specialized Access applicant with the US government. Existing Project Glasswing members move into this tier without having to reapply.
All 3 tiers have Claude Opus 5.5, Claude Sonnet 5.5, and Claude Mythos 5.1 available.
Anthropic’s overview of the Cyber Verification Program tiers, published October 6, 2026.
Opus 5.5 completed 34 of 50 attack tasks at the Red Team tier
Anthropic tested the tiers on CyScenarioBench, running Opus 5.5 five times on each of 10 multi-stage cyber challenges. Every task was blocked on the first prompt without the program.
Defense Access blocked 46 of 50 trials at some point. Red Team Access blocked none, and the model completed 34 of 50 tasks, in line with its unprotected 67.6% success rate.
Opus 5.5 was released on September 22, but most of the security tasks sent to it were diverted to the older Opus 4.8 instead.
Between April and July, Glasswing partners pinpointed 129,000+ verified vulnerabilities, and Anthropic’s own open-source scanning added another 5,500 through October. Over 33,000 have been rated as either critical or high severity.
The figures are based on 33 partner reports, and Anthropic expects the actual impact to be at least five times larger. In June, Glasswing added 150 more organizations.
In August, Kraken parent Payward joined Glasswing to scan its systems and open-source dependencies, Cryptopolitan said. Mythos was dark worldwide between June 12 and July 1, after a Commerce Department export ruling barred foreign access.
The smartest crypto minds already read our newsletter. Want in? Join them.

The 225% gain in Bitcoin over the three years to Sept. 23 rests on a tiny set of sessions. In an Oct. 5 note, Grayscale head of research Zach Pandl said that less than 0.5% of trading days had enough upside that excluding them trimmed cumulative return by more than half.
But the five best days alone are enough to do the damage. Remove those and Bitcoin still has a 95% gain, well under half of its total 225%.
The Nasdaq-100 keeps a 21% profit without its 15 strongest sessions
Pandl said the figures were published in Grayscale’s research series The Stack, and used spot BTC/USD prices. Missing ten sessions takes the gain to 27%, and missing 15 drops Bitcoin into an 11% loss.
The Nasdaq-100, which tracks large nonfinancial companies listed on Nasdaq, posted a 109% return over the same period. Without the 15 best days, the index was still up 21%, a sign its returns were more evenly distributed.
Before the days were taken out, Bitcoin’s 225% return was more than double the 109% return of the Nasdaq-100.
After those days go, Bitcoin’s advantage over the index turns around. The coin ends with an 11% loss compared to the Nasdaq-100’s 21% gain.
Pandl lays out the result in terms of opportunity cost. For an asset as volatile as Bitcoin, not being in the market is a risk in itself, he wrote, even if risk is usually measured by the chance of losing money.
“Investors waiting for volatility to subside or the outlook to become clearer may find that much of the repricing has already occurred,” Pandl wrote.
Pandl backs steady exposure over market timing
Those sessions can’t be reliably spotted in advance, Grayscale says, so an investor only catches them by already holding the coin. For long-term investors seeking capital gains, its advice is to maintain steady exposure and not to try to time the market.
Pandl draws three lessons from the uneven gains. They show how difficult it is to time Bitcoin exposure, the cost of being on the “wrong side” of volatility and the price of giving up long-term upside.
Grayscale’s numbers exclude management fees and expenses, and the firm says they are illustrative.
Grayscale has seen Bitcoin’s ups and downs all year. A Pandl report in February found Bitcoin trading more like high-growth software stocks than gold after the coin fell to about $60,000 on Feb. 5 from a peak above $126,000, Cryptopolitan reported.
Pandl said Bitcoin is still in its infancy compared to gold, which served as money for thousands of years.
As Cryptopolitan reported in June, Grayscale had warned that the market was fragile because of concentrated buying by a small group of digital asset treasury firms. Pandl estimated that if 2% of the $110 trillion generational wealth transfer went into crypto, that would bring in $2.2 trillion of new demand.
The smartest crypto minds already read our newsletter. Want in? Join them.

Open interest (OI) on onchain real-world assets (RWA) perpetual futures hit new highs of $5.78 billion, according to the latest data from DefiLlama. RWA OI has grown from near zero a year ago and a staggering 25.6x increase year to date. When looking at the composition of this market, most of that OI resides on one platform.
trade[XYZ] holds $3.866 billion of the total or roughly 67% of the market as of this writing. Variational comes in second with $968.17 million or 17%. Between them, the two platforms account for about 84% of all open positions in the sector.
Traders Are Holding Stock and Commodity Exposure Onchain
RWA perps are perpetual futures contracts that track traditional asset classes like equities, stock indices, commodities and currencies. They trade on decentralized perpetual exchanges, have no expiry date and usually settle in stablecoins with leverage being standard.
Open interest calculates the total value of all contracts that are still open at a given moment. This is unlike volume, which counts every trade over a period, including quick in-and-out trades. A rising open interest is indicative of traders building positions and holding them.
It’s important to note that a trader who opens a position on an Nvidia perp owns zero Nvidia shares. The contract simply tracks the stock’s price and that is the full extent of the exposure. The $5.78 billion in open interest is separate from tokenized Treasuries and stocks, which are usually backed by assets held by an issuer or custodian.
The Market Ran Through a Single Platform for Most of the Year
At the start of the year, trade [XYZ] was the only platform in the category with other platforms offering RWA perps emerging around April. QFEX now holds $274.61 million and GMTrade has $189.72 million, while Lighter sits just above $100 million. Ondo, Extended, Entropy, edgeX and RISEx each hold under $80 million.
Variational is the one newer entrant that has pulled away from that pack. Its open interest has picked up noticeably over the past few weeks and is now closing in on $1 billion. trade[XYZ] still holds four times as much.
Weekend Access is a Big Part of the Draw
Stock exchanges close at night and on weekends. Perp DEXs don’t. A trader holding stablecoins can take a leveraged position on gold or a tech stock at 3 a.m. on a Sunday without opening a brokerage account or moving funds offchain.
That access matters most for users outside the U.S., many of whom face restrictions or high costs when buying American equities through local brokers. Crypto-native traders get something too. They can hedge or speculate on macro moves without leaving the platforms where their capital already sits.
At least 10 venues now list RWA perps, yet two of them control the bulk of open positions. Variational’s climb over recent weeks is the closest thing the leader has faced to real competition so far.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Former BitMEX CEO Arthur Hayes says the AI gold rush is headed straight for a cliff, followed by a government bailout that could push crypto prices higher. He claims the AI investment boom is unsustainable and will eventually collapse, adding that multi-trillion dollars are being wasted on AI data centers.
The Maelstrom co-founder further argued that increased investment in data centers will ultimately lead to excess computing power, causing a dramatic reduction in costs.
“If you study financial history and you study every single major technological rollout, it always is overbuilt. There always is a crash, and there always is a bailout,” he told reporters.
Hayes says Bitcoin may surge past $1 million
Right now, many firms and governments are investing billions to build AI infrastructure. As opposed to what Wall Street views as a tech bubble fueled by equity funding, Hayes sees the huge investment in data center and grid build-out similar to the commercial real estate bubble fueled by credit.
He believes lenders, governments, and other private entities are backing these massive projects based on the fallacy of perpetual demand for AI.
Thus, he anticipates the massive development glut to trigger a downturn. Predicting a premature end to the tech rally, he expects AI capital spending growth to lose steam in mid-to-late 2027, before a clear stagnation sets in 2028. The core danger, he notes, is that debt markets will continue to expand while actual development stalls, creating a textbook environment for a credit crunch.
AI infrastructure boom could create a computing glut
The possibility of oversupply would not necessarily mean AI demand is going away. But Hayes’ argument is that infrastructure investment could grow much faster than the revenue from AI services. Companies may have more data center capacity than they can profitably use if computing costs fall as more facilities are brought online.
That dynamic could benefit AI developers and users in the longer term. Lower computing costs would make it cheaper to train and run increasingly capable models, potentially encouraging more companies to deploy AI agents and other automated services. However, it could also put pressure on companies that have invested heavily in expensive infrastructure while relying on continued growth in AI demand to justify those costs.
For crypto markets, Hayes sees oversupply as an ongoing cycle rather than an immediate danger. Governments could step in to help strategic AI companies and infrastructure projects if declining returns trigger defaults or stress in the credit market. Hayes thinks the resulting liquidity could eventually flow into risk assets like Bitcoin.
While anticipating a downturn in AI investment, Hayes asserts that government bailouts are inevitable because the technology has become a cornerstone of national defense and strategic dominance.
He argues that investors who brace for the inevitable bailouts will win big, much like those who capitalized on the post-2008 financial interventions. He further predicts that those state-sponsored interventions will dwarf the post-2008 liquidity injections and will potentially flood into digital assets.
“Thankfully, we have bitcoin and other crypto to soak up that excess liquidity, and so we know the asset that’s going to perform the best when the bailout comes,” he said.
Writing on Substack, he further explained that Bitcoin will begin a long-term rally just as AI infrastructure spending slows and credit continues to flow. The subsequent wave of panicked state intervention will then act as the ultimate catalyst to push the cryptocurrency past $1 million.
Hayes says tech firms will seek infrastructure payback
Hayes also argued that the very entities driving the AI infrastructure rush, including SpaceX, OpenAI, and Anthropic, have yet to turn a profit on their massive compute development. Thus, he believes that once current data center construction wraps up, these infrastructure builders will aggressively seek collection on those multi-billion-dollar computing contracts.
The optimistic counterargument hinges on AI becoming so indispensable over the next year that surging user demand will turn these cash-burning tech companies profitable, he stated. He further contended that the AI boom is already benefiting companies such as Nvidia and memory chipmakers.
Still, he argued that the key issue for investors is whether those companies are trading at reasonable valuations relative to their expected earnings.
Hayes is also launching Flop, a crypto venture focused on AI-agent payments, expected to debut in the first quarter of 2027. To which he argued there’s a market for it, since cheaper and more plentiful computing power will encourage the growth of AI agents.
The smartest crypto minds already read our newsletter. Want in? Join them.

Vera gives us the CPU performance and memory bandwidth to run enormous amounts of orchestration, code and data processing while keeping GPUs doing what they do best.
— Mike Nicolls, president of SpaceXAI, in Nvidia’s announcement
SpaceX is not alone off the planet
According to SpaceX, the Starmind satellites would harness solar energy in sun-synchronous orbit and let off heat into space. Each satellite has a peak processing capacity of 250 kilowatts.
A similar experiment is being conducted by Google as reported by NPR. The prototype of the Project Suncatcher carried four TPUs to see how the chips with AI capability would withstand the radiation and the extreme temperatures of outer space.
SpaceX and Google are not the only companies venturing in this direction. Starcloud has successfully launched an Nvidia H100 into orbit. This makes orbital AI appear less like a science experiment and more like a budding competition for computing power.
SpaceX $40B Nvidia Chip Deal: Financing, Vera Specs and the Orbital AI Race
The economics and physics are still open
Gartner projects that worldwide spending on AI is likely to reach $2.7 trillion in 2026, reflecting a 49.5% increase, which will be attributed to infrastructure costs.
Currently, the economics of placing that infrastructure into orbit are still very complicated, as BCG estimates that orbital data facilities will cost 2.5 to 3 times more than ground facilities.
Cooling is another issue. Brookings has estimated that one orbital data center would require 2.15 million square feet of radiators.
However, heat is not the only engineering problem. Chips in space must also withstand radiation. This has led semiconductor companies to put more effort into designing chips that are radiation-hardened (rad-hard).
We have seen a major uptick in activity for rad-hard fab processes.
— Dana Neustadter, senior director at Synopsys, via Semiconductor Engineering
Radiation, temperature fluctuations, and complicated repairs all make space computing much more difficult than using the same equipment on Earth.
What the deal does to the field
The financing strengthens Nvidia’s position while making the race more expensive for smaller rivals. Reuters cites Morgan Stanley’s estimate that AI infrastructure will need $1.5 trillion in outside financing by 2028. Those pressures are already showing up as power constraints complicate the broader AI buildout.
SpaceX is trying to raise around $40 billion with the help of Apollo Global Management to acquire AI chips from Nvidia, as per Financial Times. The importance of this deal lies in the fact that these chips can be used to power SpaceXAI’s data centers on the planet, as well as its planned AI infrastructure in orbit.
Media reports suggest that the financing will consist of approximately $10 billion in bank loans and $30 billion in the form of investment-grade debt. The transaction is expected to conclude in 2027, with PIMCO being one of the lenders that have been approached. Following the news, SpaceX stock dropped by 1% after-hours trading, while Nvidia gained 0.5%.
Why the financing points past Earth
Nvidia says SpaceXAI will use Vera CPUs and its Vera Rubin platform as Grok’s infrastructure grows toward gigawatts of computing capacity. The first-generation Starmind satellite is also set to use an optimized Vera Rubin NVL72 system.
Vera is Nvidia’s first CPU built for AI agents. It has 88 Olympus cores and up to 1.2 TB/s of memory bandwidth. Nvidia says it can complete some tasks up to 1.8 times faster than x86 processors.
Vera gives us the CPU performance and memory bandwidth to run enormous amounts of orchestration, code and data processing while keeping GPUs doing what they do best.
— Mike Nicolls, president of SpaceXAI, in Nvidia’s announcement
SpaceX is not alone off the planet
According to SpaceX, the Starmind satellites would harness solar energy in sun-synchronous orbit and let off heat into space. Each satellite has a peak processing capacity of 250 kilowatts.
A similar experiment is being conducted by Google as reported by NPR. The prototype of the Project Suncatcher carried four TPUs to see how the chips with AI capability would withstand the radiation and the extreme temperatures of outer space.
SpaceX and Google are not the only companies venturing in this direction. Starcloud has successfully launched an Nvidia H100 into orbit. This makes orbital AI appear less like a science experiment and more like a budding competition for computing power.
SpaceX $40B Nvidia Chip Deal: Financing, Vera Specs and the Orbital AI Race
The economics and physics are still open
Gartner projects that worldwide spending on AI is likely to reach $2.7 trillion in 2026, reflecting a 49.5% increase, which will be attributed to infrastructure costs.
Currently, the economics of placing that infrastructure into orbit are still very complicated, as BCG estimates that orbital data facilities will cost 2.5 to 3 times more than ground facilities.
Cooling is another issue. Brookings has estimated that one orbital data center would require 2.15 million square feet of radiators.
However, heat is not the only engineering problem. Chips in space must also withstand radiation. This has led semiconductor companies to put more effort into designing chips that are radiation-hardened (rad-hard).
We have seen a major uptick in activity for rad-hard fab processes.
— Dana Neustadter, senior director at Synopsys, via Semiconductor Engineering
Radiation, temperature fluctuations, and complicated repairs all make space computing much more difficult than using the same equipment on Earth.
What the deal does to the field
The financing strengthens Nvidia’s position while making the race more expensive for smaller rivals. Reuters cites Morgan Stanley’s estimate that AI infrastructure will need $1.5 trillion in outside financing by 2028. Those pressures are already showing up as power constraints complicate the broader AI buildout.
The smartest crypto minds already read our newsletter. Want in? Join them.

On October 5, the Ethereum Economic Zone (EEZ) transferred 0.001 ETH from Ethereum’s base chain to a layer-2 network in a single atomic transaction.
The run was a demo. EEZ has not shipped a product and nothing has changed in the protocol of Ethereum.
Two-way calls, nested calls across chains, and real-time proving are still on the roadmap as work running through 2027. The core software of EEZ is still in its infancy.
Etherscan logs the transfer in block 26,127,444 for an $0.80 fee
Etherscan shows the transaction was confirmed in block 26,127,444 at 16:44 UTC.
The 0.001 ETH, around $2.70, was sent to a contract called “EEZ: Cross Chain Proxy.” The fee was about $0.80. And in the same all-or-nothing move, a rollup’s state changed as well.
If any linked action fails, all actions are rolled back. So the L1 transfer and L2 update either both happen together or none happen at all.
The transaction was posted on X by Eduardo Antuña Díez, a core engineer on the EEZ project. “Atomic synchronous composability is no longer a promise,” he wrote.
He dubbed it “the first atomic cross-chain L1->L2 transaction ever,” adding: “And this is only the beginning for EEZ.”
Prior to the mainnet run, the team conducted an audit, tested live blob encoding and ran the system against CoW Swap and Uniswap v4.
The EEZ cross-chain transaction on Etherscan, confirmed in block 26127444 on October 5, 2026.
Gnosis and Zisk aim to rejoin liquidity split across rollups
EEZ is run by Gnosis and Zisk and is bolstered by the Ethereum Foundation. The project started in early 2026 and in March sketched a design for rollups and mainnet to call each other in a single transaction, with no bridge in between.
Its target is liquidity split across Ethereum’s many L2s. According to Gnosis co-founder Friederike Ernst, protocols would have to keep separate deployments on each L2 without synchronous composability.
Cryptopolitan reported in November 2025 that the Ethereum Foundation is also working towards the same goal with its Ethereum Interop Layer, which it said would make Ethereum “feel like one chain again.” Built on ERC-4337 account abstraction, it lets users sign once for a cross-chain transaction.
In February, Vitalik Buterin said that the “original vision of L2s and their role in Ethereum no longer makes sense,” Cryptopolitan reported. He said L2 progress, “secondarily, on interop,” had been “far slower and more difficult than originally expected.”
The reaction on X was hot. DeFi protocol Hydration co-founder Jakub Gregus ranked the demo among the most important milestones of crypto and said it could directly benefit ETH. “I told you couple months ago I am becoming increasingly bullish on Ethereum as technology,” he wrote.
Another user said that EEZ “one-shotted Ethereum’s biggest problem.” But a single 0.001 ETH call in one direction leaves most of the work for the 2027 roadmap.
If you're reading this, you’re already ahead. Stay there with our newsletter.

Fair Isaac, the credit-scoring firm behind FICO, is reducing its workforce by almost 15% as part of a bigger restructuring focused on AI. The move could affect roughly 570 employees. FICO told Reuters the changes should help it “operate and bring innovations to market faster.”
The moment is remarkable. Corporations are investing a lot in AI technology but at the same time we cannot observe any signs that it causes mass displacement in the labor market.
What FICO is cutting and why its shares are under pressure
The number of employees of FICO amounted to 3,811 at the end of September 2025. It projects approximately $27 million in pre-tax restructuring charges in the fourth quarter of financial year 2026, mostly for severance. The company aims to complete the program by Q3 of financial year 2027. According to Reuters, FICO’s stock has dropped by approximately 58% this year.
Part of the stress is also due to the modifications in the process of determining mortgage credit scores. The Federal Housing Finance Agency has recently permitted lenders to employ the VantageScore in addition to FICO when issuing mortgage loans to customers who wish to sell their mortgages to Fannie Mae or Freddie Mac. Thus, FICO will now face tougher competition in a segment where it has always been the leader.
A corporate trend increasingly called “restructuring”
FICO is not the only company to do so. Several companies in technology, banking, and other industries are laying off, while making investments in automation and AI.
Reuters reveals that the trend is growing among major employers, and Programs.com estimates that over 170,000 positions are affected by layoffs connected to AI by the end of 2026. The numbers are striking but do not state how many positions were eliminated due to the implementation of AI.
An analysis from Nexford finds that more firms are classifying job cuts related to artificial intelligence as part of broader “restructuring” plans. FICO is no exception to this trend. The company stated its “simplified structure” would enable it to “operate and bring innovations to market faster,” while generating higher value for its customers.”
However, this kind of wording blurs the picture. This makes it difficult to ascertain which portion of the cuts can be attributed to the implementation of automation technologies and which can be attributed to standard cost-reduction measures.
FICO AI restructuring: How its job cuts compare with Microsoft, HSBC and Amazon
The case that AI is finally paying off
There is also a stronger business case for keeping the spending going.
Boston Consulting Group found that nearly half of the companies it surveyed are now generating measurable value from AI.
At the same time, Gartner expects worldwide AI spending to reach about $2.7 trillion in 2026, up 49.5% from last year, with infrastructure taking the biggest share.
Why the layoffs still don’t prove an “AI jobs apocalypse”
The contradiction is that highly visible corporate cuts have not yet translated into clear evidence of mass AI-driven unemployment.
Stanford research points to a generally weaker labor market, but not one where AI clearly stands out as the main cause. The pressure appears more concentrated among younger workers and entry-level roles.
Economists Alex Imas and Jacob Schaal reach a similar conclusion. AI may already be weighing on junior hiring, but broader disruption has yet to appear clearly in the data.
Cryptopolitan previously reported on economists pushing back against Silicon Valley’s more dramatic job-loss forecasts.
FICO’s restructuring matters. But for now, it is still one piece of a much bigger debate over whether AI is eliminating jobs or simply changing where companies spend and hire.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.