Daily News Coverage · October 20, 2026
The Big Headline: Nscale-Anthropic $45B, Samsung-TSMC HBM, Ransomware Negotiator Arrest, Tenstorrent $1.5B
Good morning, folks! Nscale is in talks with Anthropic for a $45 billion cloud computing deal at its West Virginia data center. Jensen Huang says the AI sell-off is overblown and points to growth beyond OpenAI and Anthropic. Samsung will use TSMC's 3nm process for Nvidia's custom HBM base dies, and TSMC shares rose over 3% on the news. Federal authorities arrest a Canadian ransomware negotiator suspected of helping the ShinyHunters extortion group. The n0n maritime extortion incident appears to trace back to a single vendor, CyberLogitec. Tenstorrent raises $1.5 billion at a $6.2 billion valuation. Labelbox closes over $100 million. Parallel Loop raises $6.6 million for enterprise AI integration. Illinois Governor Pritzker creates an AI cabinet to explore regulation. And Nvidia wins a CPU order from AI startup Modal. Here's today's briefing from The CODEW's The Newsroom.
Nscale in Talks with Anthropic for $45 Billion Cloud Computing Deal in West Virginia
The deal would be located at the second facility of Nscale's Monarch data center campus and provide 460 megawatts of compute powered by Nvidia's Vera Rubin chips.
What happened: Nscale Limited is in talks with Anthropic for a compute supply agreement valued at approximately $45 billion, according to people familiar with the negotiations. The deal would be located at the second facility of Nscale's Monarch data center campus in West Virginia, relying on Nvidia's advanced Vera Rubin AI chips to provide 460 megawatts of compute capacity. The transaction could be finalized within days and would significantly boost Nscale's IPO prospects. The company is seeking a valuation of approximately $30 billion when it lists. Nscale raised $3.36 billion in a pre-IPO convertible round led by Third Point in late September, with a $1 billion commitment from Nvidia expected in mid-November.
Key numbers/companies: Nscale · Anthropic · $45 billion · West Virginia · Monarch data center · 460 MW · Nvidia Vera Rubin · $30 billion IPO valuation · Third Point.
Why it matters: The deal would be one of the largest AI compute contracts ever signed and would deepen Anthropic's reliance on Nvidia-powered infrastructure outside of its existing Google TPU arrangements. For Nscale, the agreement would provide anchor demand for its West Virginia campus and strengthen its position as a credible alternative to CoreWeave and Lambda in the AI cloud market.
What's next: Whether the deal is finalized in the coming days and how it affects Nscale's IPO timeline.
Sources: Techmeme
Jensen Huang Says AI Sell-Off Over OpenAI and Anthropic Revenue Is Overblown
The Nvidia CEO points to xAI's Grok, Meta's Muse, and Google's Gemini as evidence that AI demand extends well beyond the two leading labs.
What happened: Nvidia CEO Jensen Huang said on a podcast with venture capitalist Brad Gerstner that fears over OpenAI and Anthropic revenue are overblown after AI stocks sold off sharply in October. His argument rests on three points: there are more than two frontier labs, with xAI's Grok, Meta's Muse agent, and Alphabet's Gemini all "doing great" and growing; Anthropic's Q3 revenue of roughly $16.8 billion is a remarkable achievement given it generated only $4.6 billion in all of 2025; and the AI build-out is bigger than any single company. The sell-off was triggered in part by a Financial Times report that OpenAI had reached $50 billion in annualized recurring revenue, well short of the $70 billion figure leaked in September, and by Anthropic's own investor presentation showing a Q3 annualized revenue figure of $67.2 billion versus a leaked $65 billion ARR at the end of July.
Key numbers/companies: Nvidia · Jensen Huang · OpenAI · Anthropic · $50 billion ARR · $70 billion leak · $16.8 billion Q3 revenue · $67.2 billion annualized · Grok · Muse · Gemini.
Why it matters: Huang's intervention is significant because Nvidia is the single largest beneficiary of AI infrastructure spending. If he is right that demand is broader than OpenAI and Anthropic, the sell-off may be overdone. If he is wrong, the correction has further to run. His comments also reveal a strategic shift in Nvidia's messaging: the company is no longer positioning itself as dependent on the two leading labs.
What's next: Whether the market accepts Huang's framing and how OpenAI and Anthropic revenue trends evolve in Q4.
Sources: Mitrade; PrimeXBT; The Motley Fool; Axios.
Samsung to Use TSMC's 3nm Process for Nvidia's Custom HBM Base Dies, TSMC Shares Rise 3%
The move marks the first collaboration between the rivals at the HBM manufacturing stage and expands TSMC's role in Nvidia's AI chip supply chain beyond processors.
What happened: Samsung Electronics will use TSMC's 3-nanometer process to manufacture base dies for Nvidia's next-generation custom high-bandwidth memory chips, the Seoul Economic Daily reported. The move marks the first time Samsung has worked with TSMC at the HBM manufacturing stage. Nvidia has asked Samsung, SK hynix, and Micron to use TSMC's 3nm process for the base dies and a sixth-generation DRAM process for the memory chips' core dies. Nvidia unveiled its custom memory technology, called Nvidia HBM, in August. The design integrates Nvidia's custom memory controller into the base die and aims to standardize the architecture across multiple memory suppliers. Moving the memory controller from the GPU to the base die frees up space on the GPU for additional computing circuits. Samsung will continue manufacturing the memory layers while relying on TSMC's process for the base die. TSMC shares rose more than 3% on the news.
Key numbers/companies: Samsung · TSMC · Nvidia · 3nm · HBM · base die · SK hynix · Micron · TSMC +3%.
Why it matters: The arrangement is a strategic reversal for Samsung, which has promoted its one-stop-shop model for years. By ceding part of the HBM manufacturing process to its biggest rival, Samsung is acknowledging that TSMC's yield and process technology are critical to meeting Nvidia's standards. For TSMC, the deal expands its role in Nvidia's AI supply chain beyond processor manufacturing and gives it leverage in the memory market.
What's next: Whether Samsung's own foundry qualifies to produce Nvidia HBM base dies if TSMC capacity becomes constrained.
Sources: Taiwan News; Reuters; Seoul Economic Daily; Investing.com.
FBI Arrests Canadian Ransomware Negotiator Suspected of Helping ShinyHunters Extortion Group
Edward Dubrovsky allegedly knew ShinyHunters had stolen sensitive data from FBI servers and did not inform the bureau before the group threatened to publish personnel files.
What happened: Federal authorities arrested Edward Dubrovsky, a 54-year-old Canadian cybersecurity executive involved in ransomware negotiation, on charges related to assisting the ShinyHunters hacking group. ShinyHunters is known for stealing sensitive data, including from FBI servers, and extorting victims. Dubrovsky allegedly had knowledge of these breaches and did not inform the FBI prior to public threats by ShinyHunters. According to KrebsOnSecurity, Signal messages showed Dubrovsky had frequent conversations with a Dutch hacker and at least one executive at a large U.S.-based consulting firm, and these executives often knew details about new ShinyHunters victims before the victims themselves learned of the breach. The case is part of a broader FBI investigation centralized in Texas, with other arrests and ongoing inquiries into ransomware negotiation firms.
Key numbers/companies: Edward Dubrovsky · ShinyHunters · FBI · KrebsOnSecurity · ransomware negotiation · FBI personnel data.
Why it matters: The arrest raises serious questions about conflicts of interest in the ransomware negotiation industry. If negotiators are colluding with the groups they are supposed to be defending against, the entire incident response model is compromised. The case also suggests that ShinyHunters had inside knowledge of victim breaches before public disclosure.
What's next: Whether more arrests follow and how ransomware negotiation firms respond to the investigation.
Sources: KrebsOnSecurity; Threat Radar; Cybernoz; Cyberworldops.
n0n Maritime Extortion Incident Likely Traces Back to a Single Vendor, CyberLogitec
At least 21 of the 38 maritime organizations listed by n0n are documented CyberLogitec customers, suggesting a single compromised support archive rather than 38 separate breaches.
What happened: The n0n cyber extortion group listed 38 maritime organizations in 18 countries on its dark web leak site on October 12, with ransom demands ranging from $400,000 to $400 million per company and a common deadline of October 19, 2026, 13:27 UTC. Cydome's research team assessed that the incident is likely linked to data from CyberLogitec, a Korean company that provides software and IT services to maritime companies. At least 21 of the 38 companies are publicly documented CyberLogitec customers, and seven sample file names include the vendor's name or one of its product names. The stolen data likely comes from a single support and project archive linked to CyberLogitec, split into separate listings per customer, with file dates from 2009 to August 2026. The samples include credentials, personal data, EDI traffic, and system documentation. Being named on the list does not necessarily mean an organization's own network was breached.
Key numbers/companies: n0n · CyberLogitec · 38 maritime organizations · 18 countries · $400K-$400M ransom · 21 documented customers · Cydome · October 19 deadline.
Why it matters: The incident reframes the n0n campaign from 38 separate breaches to a single supply chain compromise. If confirmed, it highlights the systemic risk that software vendors pose to their customers — a single compromised archive can expose dozens of organizations simultaneously. It also demonstrates the value of shared infrastructure to attackers.
What's next: Whether n0n follows through on its leak threat and how CyberLogitec responds.
Sources: Hellenic Shipping News; Cydome; IndexBox; DeXpose.
Tenstorrent Raises $1.5 Billion at $6.2 Billion Valuation to Challenge Nvidia in AI Servers and Chip IP
The Series E round was led by Trace Capital and Innovation X, with participation from Fidelity, T. Rowe Price, SanDisk, ARK Invest, and Morgan Stanley.
What happened: Tenstorrent has raised $1.5 billion at a $6.2 billion post-money valuation, giving the AI chipmaker fresh capital to scale production of its Galaxy servers and expand its processor licensing business as it competes for workloads dominated by Nvidia. The Series E round was led by Trace Capital and Innovation X, with participation from Fidelity Management & Research Company, T. Rowe Price, SanDisk, ARK Invest, Morgan Stanley, and existing investors. Tenstorrent plans to use the funding to expand Galaxy server manufacturing and commercial deployments, develop new generations of its AI and RISC-V CPU intellectual property, and hire globally. The company has named several commercial deployments: Virtu Financial has deployed Galaxy servers on-premises for AI agents supporting trading and operations; Equinix offers Galaxy systems through its Distributed AI Hub; and ai& in Japan has deployed more than 120 Galaxy systems as part of its sovereign AI infrastructure. The new valuation is more than twice the roughly $2.7 billion implied post-money value of its December 2024 financing.
Key numbers/companies: Tenstorrent · $1.5 billion · $6.2 billion post-money · Trace Capital · Innovation X · Fidelity · T. Rowe Price · SanDisk · ARK Invest · Morgan Stanley · Galaxy servers · RISC-V · Virtu Financial · Equinix · ai&.
Why it matters: The raise signals continued investor appetite for Nvidia alternatives, even as the AI trade cools. Tenstorrent's RISC-V CPU technology and Galaxy servers position it as a potential on-premises and sovereign AI alternative to Nvidia's GPU-dominated stack. The $6.2 billion valuation is ambitious but reflects the premium investors are placing on CEO Jim Keller's track record and the scarcity of credible alternatives.
What's next: Whether Tenstorrent can scale commercial deployments and demonstrate predictable revenue growth.
Sources: TechTarget; Yahoo Finance; Tenstorrent.
Labelbox Raises Over $100 Million Series E at $1.25 Billion Pre-Money Valuation for RL Infrastructure
The company serves over 400 customers with reinforcement learning environments and a managed agent platform, and says it has been profitable in 2026 with revenue up more than fivefold.
What happened: Labelbox has completed a Series E funding round of more than $100 million, including equity and a credit facility, at a pre-money valuation of $1.25 billion. The round was led by Infinitum Partners, with participation from Wipro Ventures, a16z, B Capital, Tau Ventures, Snowpoint Ventures, and others. The San Francisco-based company provides reinforcement learning environments and evaluation infrastructure (Horizon) for frontier AI labs and a managed agent platform (Recursion Managed Agents) for enterprises. It serves more than 400 customers and says it has been profitable in 2026, with revenue up more than fivefold year-over-year. The funding will be used for R&D and team expansion across both product lines.
Key numbers/companies: Labelbox · $100 million+ · $1.25 billion pre-money · Infinitum Partners · Wipro Ventures · a16z · B Capital · 400+ customers · Horizon · Recursion Managed Agents.
Why it matters: Labelbox sits at the intersection of two growing markets: reinforcement learning infrastructure for frontier labs and managed agent platforms for enterprises. The company's profitability and revenue growth stand out in a sector where many AI startups are burning cash. The raise suggests investors see durable value in the data and evaluation layer that underpins model development.
What's next: How Labelbox deploys the capital and whether it pursues an IPO.
Sources: Investing.com
Parallel Loop Raises $6.6 Million Seed to Embed AI Engineers Inside Enterprises
The Singapore-based startup targets the gap between AI adoption and operational integration, with nearly nine in 10 AI-using Singapore firms still relying on off-the-shelf tools.
What happened: Singapore-based enterprise AI startup Parallel Loop has raised $6.6 million in seed funding led by Bengaluru-based venture capital firm 3one4 Capital. The round attracted participation from Germany's Picus Capital, fintech investor Flourish Ventures, and Economic Development Innovations Singapore (EDIS), alongside angels including Paytm founder Vijay Shekhar Sharma, CRED founder Kunal Shah, and Philip Yeo, former chairman of Singapore's Economic Development Board. The company did not disclose its valuation. Parallel Loop combines a proprietary AI software platform with engineers embedded within client organizations to connect existing technology systems, automate workflows, and develop applications. It formally launched on October 14 and serves more than 10 enterprise customers, including Havells, Licious, and Tokio Marine Safety Insurance. The startup has a 30-member team across Singapore, Bengaluru, Bangkok, London, and Dubai.
Key numbers/companies: Parallel Loop · $6.6 million seed · 3one4 Capital · Picus Capital · Flourish Ventures · EDIS · Vijay Shekhar Sharma · Kunal Shah · Philip Yeo · Havells · Licious · Tokio Marine.
Why it matters: The raise reflects growing investor interest in the services layer of enterprise AI — the people and processes required to make AI actually work inside complex organizations. Singapore's Digital Economy Report 2026 found that 23.4% of SMEs and 70.4% of larger enterprises have adopted AI, but nearly nine in 10 AI-adopting enterprises rely on off-the-shelf tools and only 28.3% customize them. Parallel Loop's forward-deployed engineering model addresses that gap.
What's next: How quickly Parallel Loop scales its embedded engineering model across Asia-Pacific.
Sources: Entrepreneur Asia Pacific
Illinois Governor Pritzker Creates AI Cabinet to Explore Regulation Through 2027
The cabinet will assess AI risks to critical infrastructure, consider tying data center incentives to safety standards, and evaluate whether existing laws can hold AI companies accountable.
What happened: Illinois Governor JB Pritzker signed an executive order establishing an AI Cabinet made up of experts and representatives from several state agencies. The unpaid group will formulate policy proposals through 2027, analyzing and assessing "the various risks posed by AI to the safety of Illinois residents and the critical infrastructure residents rely on." The cabinet will consider regulations tying data center incentives or operational approval to a company's ability to meet safety, accountability, and transparency standards; address strain on the energy grid from data centers; evaluate how state procurement can incorporate AI safety measures; and assess whether existing legal options are sufficient to hold AI companies accountable. Pritzker signed Senate Bill 315 in July, the Artificial Intelligence Safety Measures Act, which increases transparency and accountability requirements for the largest AI models — those generating more than $500 million in annual revenue and trained using massive computing power. It made Illinois the first state to require annual audits of AI, effective 2028. Pritzker has called on Congress to impose national AI safety standards, saying the technology is "dangerous on the level of nuclear weapons."
Key numbers/companies: Illinois · JB Pritzker · AI Cabinet · Senate Bill 315 · $500 million revenue threshold · annual AI audits · 2028 effective date.
Why it matters: Illinois is moving aggressively on AI regulation while the federal government takes a hands-off approach. The AI Cabinet's scope — from data center energy strain to procurement standards to legal accountability — signals that states are preparing to regulate AI across multiple dimensions, not just model safety. The state's audit requirement, the first in the nation, could become a template for others.
What's next: The cabinet's policy recommendations and whether other states follow Illinois's audit model.
Sources: Insurance Journal; Capitol News Illinois.
Nvidia Wins AI Startup Modal as Customer for Its Vera CPU, Challenging Intel and AMD
OpenAI, Anthropic, Oracle, and Perplexity are also planning to adopt Vera, expanding Nvidia's AI infrastructure footprint beyond GPUs.
What happened: Nvidia has secured AI startup Modal as a customer for its Vera CPU, according to Cnyes. Modal's annualized revenue is estimated to exceed $1 billion. Beyond Modal, Nvidia has disclosed that OpenAI, Anthropic, Oracle, and Perplexity are also planning to adopt Vera. The wins expand Nvidia's footprint from GPUs into the CPU market, where Intel and AMD have historically dominated.
Key numbers/companies: Nvidia · Vera CPU · Modal · $1 billion+ annualized revenue · OpenAI · Anthropic · Oracle · Perplexity · Intel · AMD.
Why it matters: Nvidia's move into CPUs represents a significant expansion of its competitive territory. If Vera gains traction with major AI labs and cloud providers, it could erode Intel's and AMD's share of the data center CPU market. The wins also reinforce Nvidia's strategy of controlling more of the AI infrastructure stack — from GPUs to CPUs to networking.
What's next: Whether Vera adoption accelerates and how Intel and AMD respond.
Sources: Cnyes
Why It Matters
Nscale's $45 billion Anthropic deal would be the largest AI compute contract of the year. It signals that Anthropic is diversifying its infrastructure beyond Google TPUs and that Nvidia-powered cloud providers can compete for frontier lab workloads. For Nscale, the deal would validate its West Virginia campus and strengthen its IPO case. For the broader market, it shows that demand for AI compute remains strong despite the equity sell-off.
Jensen Huang's pushback on the sell-off is a strategic defense of Nvidia's customer base. By pointing to xAI, Meta, and Google as growth drivers beyond OpenAI and Anthropic, Huang is trying to decouple Nvidia's outlook from the revenue trajectories of two companies. Whether the market accepts that framing will determine whether the AI trade stabilizes or continues to correct.
Samsung's TSMC concession is a milestone in the AI supply chain. The two rivals are now cooperating at the HBM manufacturing stage because Nvidia's standards require it. TSMC's 3nm process and yield have become the quality gatekeeper for Nvidia's custom memory, giving TSMC leverage it did not previously have in the memory market.
The ransomware negotiator arrest exposes a structural conflict of interest. If the people hired to protect victims are colluding with the attackers, the entire incident response model is compromised. The ShinyHunters case suggests that some negotiators knew about breaches before the victims did — a finding that should prompt a re-evaluation of how ransomware negotiations are governed.
The n0n maritime incident is a supply chain story, not 38 separate breaches. If Cydome's assessment is correct, a single compromised archive at CyberLogitec exposed dozens of maritime organizations. That reframes the risk: the threat is not just individual companies' security posture, but the security of the vendors they depend on.
The CODEW Take
Nscale's $45 billion Anthropic deal is the clearest signal yet that the AI compute market is not cooling — it is consolidating. The equity sell-off is real, but the contract flow tells a different story. Anthropic is locking in 460 megawatts of Nvidia-powered capacity in West Virginia. Nscale is using that anchor demand to support a $30 billion IPO. The companies that control compute capacity are still in a strong position. The ones that financed at 10%+ yields are the ones who should worry.
Jensen Huang's podcast appearance was a masterclass in narrative management. He did not deny that OpenAI and Anthropic revenue growth has slowed. He reframed the question: AI demand is bigger than two companies. xAI's Grok, Meta's Muse, Google's Gemini — all growing. Nvidia is not dependent on any single lab. That may be true, but it is also convenient. The market will decide whether Huang is right or whether he is defending his stock.
Samsung's decision to use TSMC's 3nm process is a humbling admission. For years, Samsung promoted its one-stop-shop model as a competitive advantage. Now it is ceding the base die — the most technologically demanding part of Nvidia's custom HBM — to its biggest rival. The reason is simple: TSMC's yield exceeds 90%, and Samsung's does not. Nvidia's standards are now the benchmark, and TSMC is the only foundry that meets them at scale for this product.
The ransomware negotiator arrest is the most important cybersecurity story of the week. The FBI arrested Edward Dubrovsky on charges related to assisting ShinyHunters. According to KrebsOnSecurity, Signal messages showed he knew about the FBI data breach before the public did — and never told the bureau. If negotiators are colluding with attackers, the trust model that underpins the entire incident response industry is broken. Every CISO should be asking their negotiation firm: who exactly are you working for?
The n0n maritime incident is a case study in supply chain concentration risk. Thirty-eight maritime organizations appeared on n0n's leak site, but Cydome's analysis suggests the data came from a single CyberLogitec archive. That means the victims may not have been breached individually — they were exposed through a shared vendor. The lesson is not new, but it is increasingly urgent: in a world where every company depends on software vendors, the weakest link in the supply chain is the entry point for everyone else.
What to Watch
- Nscale-Anthropic deal close: Whether the $45 billion compute agreement is finalized in the coming days.
- Nvidia GTC Berlin (Oct 20-22): Jensen Huang's keynote and European AI infrastructure announcements.
- AI credit markets: Whether the sell-off stabilizes or deepens after Huang's comments.
- Samsung-TSMC HBM production: When Samsung begins producing Nvidia HBM base dies on TSMC's 3nm process.
- ShinyHunters investigation: Whether more arrests follow and how ransomware negotiation firms respond.
- n0n leak follow-through: Whether the group publishes data and how CyberLogitec responds.
- Tenstorrent execution: Whether the company can scale Galaxy server deployments and demonstrate revenue growth.
- Labelbox and Parallel Loop capital deployment: How the companies use their new funding to expand.
- Illinois AI Cabinet recommendations: The policy proposals that emerge through 2027.
- Nvidia Vera CPU adoption: Whether Modal, OpenAI, Anthropic, Oracle, and Perplexity follow through on planned deployments.
Editorial Note
The CODEW Daily News Coverage tracks the most important technology developments of the day, with a focus on AI, enterprise software, cloud computing, semiconductors, cybersecurity, startups, and digital infrastructure. This edition is a pure news aggregation and coverage page — what happened — while the Daily Tech Briefing provides the deeper strategic interpretation.
Coverage is based on company announcements, public disclosures, industry reporting, and other publicly available information. Reported figures and sourced-but-unconfirmed details are noted as such.
Reviewed by Erwin Castro
on
Tuesday, October 20, 2026
Rating:
No comments: