Nvidia is in advanced talks to become an anchor investor in Anthropic's initial public offering, potentially committing up to 10 billion dollars. This move would help Anthropic raise nearly 100 billion dollars at a valuation of approximately 2 trillion dollars, possibly making it the largest stock offering in history. The company aims to complete this IPO before the November U.S. midterm elections.
Anthropic plans to utilize over 1 million Trainium2 chips from Amazon Web Services for its AI model, Claude. Additionally, the company has committed to purchasing 30 billion dollars of Microsoft Azure computing capacity powered by Nvidia chips. Nvidia expects this partnership to drive a significant portion of its business next year as it seeks to dominate the AI market.
Meanwhile, Anthropic has warned that users in Iran, Russia, and other nations misused its Claude AI for weapons research. A 154-page report updated on September 12, 2026, flagged attempts to use the tool for missile systems, kamikaze drone swarms, and biological weapons. Iran-linked actors and Houthi rebels specifically used the AI to analyze ship data and create guidance software for missiles targeting U.S. Navy ships.
In the hardware sector, Hewlett Packard Enterprise faces a 7.6 billion dollar backlog of AI orders it cannot fulfill due to memory supply constraints. This figure represents a major increase from the 3.5 billion dollar backlog reported in 2020. HPE AI revenue grew by 50 percent in the past year as the company focuses on meeting this growing customer interest.
On the software and security front, Google released TimesFM-3, a model with 330 million parameters trained on over one trillion data points to improve sales forecasting. It ranks first on three benchmarks, outperforming competitors like Amazon's Chronos-2. Netskope also reported strong results, with its AI-powered security platform seeing a 50 percent increase in sales during the second quarter.
OpenAI CEO Sam Altman addressed safety concerns in an exclusive interview, stating he would pause development if safety could not be ensured. He warned that AI beyond human control is possible and announced that the company's upcoming IPO will not happen until 2027 due to these safety concerns. Altman emphasized that no gamble with humanity is acceptable.
Key Takeaways
- Nvidia is considering investing up to 10 billion dollars as an anchor investor in Anthropic's IPO.
- Anthropic plans to raise nearly 100 billion dollars at a valuation of about 2 trillion dollars.
- Anthropic expects to finish its IPO before the November U.S. midterm elections.
- Anthropic plans to use more than 1 million Trainium2 chips from Amazon Web Services for Claude.
- Anthropic committed to purchasing 30 billion dollars of Microsoft Azure computing capacity powered by Nvidia chips.
- Anthropic flagged misuse of Claude for weapons research by users in Iran, Russia, and other countries.
- Hewlett Packard Enterprise has a 7.6 billion dollar AI backlog waiting for memory supply.
- HPE AI revenue grew by 50 percent in the past year amid high demand.
- Google released TimesFM-3, which ranks first on three sales forecasting benchmarks.
- OpenAI CEO Sam Altman announced the company's IPO will not happen until 2027 due to safety concerns.
Nvidia may invest 10 billion in Anthropic IPO
Nvidia is considering becoming an anchor investor in Anthropic's initial public offering. Reuters reported that Nvidia might put up to 10 billion dollars for this deal. Anthropic plans to raise nearly 100 billion dollars at a valuation of about 2 trillion dollars. This potential deal could make it the largest stock offering ever. The company expects to finish the IPO before the November U.S. midterm elections.
Nvidia weighs 10 billion stake in Anthropic IPO
Nvidia Corp is talking about investing up to 10 billion dollars in Anthropic's upcoming stock sale. This move would help Anthropic raise as much as 100 billion dollars at a 2 trillion dollar valuation. The talks are still early and could change before the November U.S. midterm elections. Anthropic plans to use more than 1 million Trainium2 chips from Amazon Web Services for its AI model Claude. The company also expects revenue to reach between 190 billion and 200 billion dollars by 2028.
Nvidia seeks 10 billion deal in Anthropic IPO
NVIDIA is in talks to invest up to 10 billion dollars as an anchor investor in Anthropic's IPO. This investment could value the AI company at 2.3 trillion dollars and make it the largest stock offering ever. The IPO is targeted to happen before the U.S. midterm elections in November 2026. NVIDIA expects this partnership to drive a significant portion of its business next year. This move aligns with NVIDIA's strategy to dominate the AI market by securing compute demand.
Anthropic seeks 10 billion Nvidia investment for IPO
Anthropic is in talks to bring Nvidia in as an anchor investor for its proposed mega-listing. The company is seeking to raise as much as 100 billion dollars, which could value the Claude maker at about 2 trillion dollars. Nvidia is considering investing up to 10 billion dollars in this deal. In November 2025, Nvidia said it would invest up to 10 billion dollars in Anthropic as part of a broader partnership. Anthropic also committed to purchasing 30 billion dollars of Microsoft Azure computing capacity powered by Nvidia chips.
Anthropic warns Iran Russia used Claude for weapons
Anthropic says users in Iran, Russia, and other countries misused its Claude AI for weapons research. The company flagged attempts to use the tool for missile systems and kamikaze drone swarms. Reports also show users tried to use Claude for biological weapons research, surveillance, and hacking. A 154-page report cited by Bloomberg listed users from Iran, Russia, China, and Yemen among those flagged. The report was updated on September 12, 2026.
Iran and Houthis used Anthropic Claude AI for weapons
Iran-linked actors and Houthi rebels used Anthropic's Claude AI to target U.S. Navy ships and build advanced weapons. They utilized the tool to analyze ship data, find communication flaws in systems like Cobham Sailor and Cisco equipment, and create guidance software for missiles. The group developed code for a guided rocket, a ballistic missile with a 2,000 km range, and a hypersonic glide vehicle. Anthropic banned the malicious accounts and shared its findings with government authorities after discovering these activities.
HPE has 7.6 billion dollar AI backlog waiting for memory
Hewlett Packard Enterprise has a 7.6 billion dollar backlog of artificial intelligence orders that it cannot fulfill yet. CEO Antonio Neri explained that the company is waiting for memory supply to catch up with the high demand from enterprise and government customers. This backlog represents a major increase from the 3.5 billion dollar figure reported in 2020. HPE AI revenue grew by 50 percent in the past year as the company focuses on meeting this growing customer interest.
New SIRF model handles industrial content risk automatically
Researchers introduced SIRF, a Spec-Internalized Risk Foundation Model designed for industrial content risk control. This model internalizes complex platform policies directly into its weights using a method called continued pretraining without needing extra human annotation. SIRF-8B-SFT achieved a 71.3 percent Black Recall score, which is 15.1 percentage points higher than the baseline model. The system operates with ultra-low latency and helps recover 20 percent more mis-penalized samples when used as a tree-model adjudication layer.
Portal26 launches Recon tool for real-time AI risk visibility
Portal26 launched a new tool called Recon to help enterprises monitor AI risk in real time. The platform allows security teams to ask plain language questions about AI usage and receive evidence-based answers regarding security, governance, and compliance. It analyzes prompts and responses to identify risks related to regulations like GDPR, CCPA, and ISO 42001. The tool automatically generates risk scorecards and remediation plans to help companies adjust their security posture as regulations change.
38 percent of B2B buyers use agentic AI for purchasing
New Deloitte research shows that 38 percent of B2B buyers are already using agentic AI in their purchasing decisions. However, only 13 percent of suppliers report that their front and back office systems are fully integrated. Nearly nine out of ten suppliers are currently upgrading their enterprise resource planning systems to support better automation. Suppliers with high digital maturity beat annual sales goals by 6.1 percent on average compared to 2.9 percent for those with lower maturity.
New AI tool creates realistic fake business data
Researchers created a new generator that builds complete fictional enterprises without using any real data. This tool takes inputs like industry type and company size to produce consistent workforces, customers, and sales records. It projects one entity graph into 66 different business product formats so every customer has a single identity across systems. The system uses a five-axis scorecard with 28 statistical checks to verify realism, and it improved scores from 60.3 to 99.1 over 23 generated companies. A second generator in the project builds relational databases directly from lists of business questions.
Google releases TimesFM-3 for better sales forecasting
Google Research launched TimesFM-3, a new AI model that predicts future trends using sales data, weather, and discount schedules. The model uses 330 million parameters and was trained on more than one trillion data points to handle multiple variables at once. Unlike earlier versions that predicted one block at a time, TimesFM-3 fills in all future time steps in a single pass to avoid compounding errors. It ranks first on three benchmarks including Gift-Eval, FEV-Bench, and Time, outperforming competitors like Amazon's Chronos-2. The tool is already used in retail, finance, manufacturing, healthcare, and science since the family launched in 2024.
Netskope AI security sales jump 50 percent in Q2
Netskope reported strong second quarter results with annual recurring revenue rising 27 percent to $899 million. The company beat its revenue guidance of $220.5 million and raised its full-year outlook. A major driver of this growth is its AI-powered security platform, which saw a 50 percent increase in sales. This success shows that the company is effectively capturing market share in the fast-growing AI security sector. Management believes the business is well positioned to continue its positive growth trajectory.
Sam Altman discusses AI safety risks and IPO plans
OpenAI CEO Sam Altman gave an exclusive interview to address growing fears about artificial intelligence destroying humanity. He stated that no gamble with humanity is acceptable and he would pause development if safety could not be ensured. Altman explained that safety standards are not ready to push AI capabilities further and warned that AI beyond human control is absolutely possible. He also announced that the company's upcoming IPO will not happen until 2027 due to these safety concerns. The interview covered national and international regulatory needs as well as the rapid development of AI models.
Sources
- Nvidia Eyes Up To $10 Billion Investment in Anthropic IPO as Claude AI Maker Mulls Nearly $100 Billion Raise at $2 Trillion Valuation: Report
- Nvidia Eyes Up To $10 Billion Investment in Anthropic IPO as Claude AI Maker Mulls Nearly $100 Billion Raise…
- NVIDIA may invest $10B in Anthropic's $2.3T IPO, aiming to secure AI dominance for the next decade.
- Anthropic in talks for up to $10 billion Nvidia investment ahead of mega IPO: Report
- Anthropic says Iran, Russia used Claude for weapons research, including missile systems and kamikaze drone projects
- Iran and Houthi rebels used Anthropic's Claude AI to target US warships and build hypersonic missiles — Houthi rebels also used the bot to code ballistic missile guidance systems
- HPE's $7.6 Billion AI Backlog Is Waiting on Memory Supply to Catch Up
- SIRF: A Spec-Internalized Risk Foundation Model for Industrial Content Risk Control
- Portal26 Launches Recon to Give Enterprises Real-Time Visibility Into AI Risk
- Deloitte: 38% of B2B Buyers Now Use Agentic AI
- Generating a Consistent Enterprise: Synthesis and Reference-Free Evaluation of Multi-System Business Data
- Google's new AI model predicts the future from sales data, weather, and discount schedules
- Netskope’s (NTSK) AI Security Push Is Already Paying Off
- Exclusive: Sam Altman addresses AI doomsday fears in new interview
Comments
Please log in to post a comment.