Nvidia Secures AI Factories While OpenAI Hires Google Executive

Investors are increasingly concerned about the high valuations of artificial intelligence stocks, including major players like Oracle and Broadcom. While AI has fueled market growth for three years, questions now arise about whether company fundamentals support these elevated prices. IBM CEO Arvind Krishna has voiced skepticism, suggesting that the billions poured into AI infrastructure, such as data centers, may not yield profitable returns. Experts advise investors to carefully review their AI holdings, drawing parallels to the dot-com bubble. Despite valuation worries, significant investments continue in AI infrastructure. Fortinet now offers its FortiGate-VM security solution for Nvidia BlueField-3, embedding networking and security directly into AI factories. This allows core security functions to run on the DPU, ensuring secure, high-performance environments for developing and training AI models without slowing critical workloads. Cisco also introduced its Integrated AI Security and Safety Framework to help organizations manage AI risks, combining system protection with ethical AI behavior. The growing complexity of AI is prompting new regulations and highlighting risks. New German and European laws, including the German Product Liability Act and the EU Artificial Intelligence Act, will extend liability for AI-driven risks beyond hardware to include software updates and data quality, particularly in high-security fields. Simultaneously, the misuse of AI is evident, with an AI-generated image spreading false claims about the Bondi Beach shooting, demonstrating the challenge of misinformation. Companies are strategically positioning themselves in the evolving AI landscape. OpenAI recently hired Kim Lee, a former top corporate development executive from Google, signaling potential plans for more acquisitions, following at least five major deals in 2025. Perplexity AI updated its iPad app with enhanced research tools and multitasking features, targeting business customers and students to boost subscription revenue. Lambda also offers a multi-cloud plan for high-performance AI infrastructure, providing access to NVIDIA GPU clusters and managed Kubernetes to mitigate risks like GPU shortages and high costs. Starling AIX, developed by Chris Kincade, is helping organizations build an "Organizational Brain" by structuring company knowledge for contextual use by large language models, aiming to preserve institutional memory. On the economic front, Ankur Crawford, an executive vice president and portfolio manager at Alger, shared her 2026 AI market predictions, anticipating an acceptable but not strong overall economic situation, with consumers showing more financial strength in the first half of 2026. Separately, in Europe, BPER Banca's Protection Certificate Memory Coupon was Italy's top-selling product in 2025, raising 490 million euros.

Key Takeaways

• Investors are increasingly concerned about high valuations of AI stocks like Oracle and Broadcom, questioning returns on massive infrastructure investments.
• IBM CEO Arvind Krishna expressed doubts about the profitability of billions invested in AI data centers.
• Cisco launched its Integrated AI Security and Safety Framework to manage AI risks, combining security and ethical safety.
• New German and EU laws, including the Product Liability Act and AI Act, will extend product liability to AI-driven risks, affecting software and data quality.
• An AI-generated image was used to spread false claims about the Bondi Beach shooting, highlighting AI's role in misinformation.
• OpenAI hired Kim Lee, a former Google corporate development executive, suggesting an increased focus on acquisitions.
• Fortinet integrated its FortiGate-VM security with Nvidia BlueField-3 to secure AI factories by offloading security functions to DPUs.
• Perplexity AI relaunched its iPad app with enhanced research tools and multitasking features, targeting business users and students.
• Lambda offers a multi-cloud AI infrastructure plan with dedicated NVIDIA GPU clusters to address GPU capacity shortages and optimize costs.
• Ankur Crawford, an executive vice president at Alger, predicts an acceptable but not strong economic situation for 2026, with stronger consumer finances in H1.

Investors Worry About High AI Stock Valuations

Artificial intelligence stocks like Oracle and Broadcom are facing growing investor concerns due to their high valuations. Companies are investing billions into AI infrastructure, such as data centers, but the return on investment is being questioned. Oracle's recent earnings showed revenue slightly below estimates and increased capital expenditure guidance. IBM CEO Arvind Krishna believes these massive investments in data centers may not yield returns. Investors are advised to carefully review AI holdings and valuations, comparing the situation to the dot-com bubble.

AI Stock Concerns Mount for Oracle and Broadcom

Artificial intelligence has boosted the stock market for three years, but investors now worry about the high valuations of AI-related stocks. Companies like Oracle and Broadcom have seen their shares surge, but analysts question if their fundamentals support these prices. Oracle benefits from AI cloud services, and Broadcom from AI chips, yet both face concerns about legacy businesses, competition, and supply chains. This has led to a split, with pure AI plays like Nvidia still rising while others stagnate. Investors should research carefully before investing in AI stocks.

AI Stock Valuations Raise Investor Concerns

Investors are increasingly worried about the high valuations of artificial intelligence stocks, including Oracle and Broadcom. Large AI companies are pouring billions into infrastructure like data centers, but questions arise about the actual return on these huge investments. Oracle's recent financial report showed revenue slightly below expectations and a significant increase in capital spending. IBM CEO Arvind Krishna stated that such massive data center investments might not be profitable. Experts suggest investors should carefully check their AI stock holdings and valuations, remembering the internet's dot-com bubble.

Cisco Launches New AI Security and Safety Framework

Cisco introduced its new Integrated AI Security and Safety Framework to help organizations manage AI risks. This framework combines AI security, which protects systems from threats, and AI safety, which ensures ethical and reliable AI behavior. Existing guidelines from MITRE ATLAS, NIST, and OWASP only cover parts of the risk landscape. Cisco's framework is unique because it integrates threats and harms, considers the AI development lifecycle, multi-agent coordination, multimodality, and audience-aware utility. This unified approach helps businesses build robust, responsible, and trustworthy AI systems.

New Laws Address AI Product Risks in High-Security Fields

New German and European laws are changing how companies handle product liability for AI-driven risks in high-security areas. The new German Product Liability Act and the EU Artificial Intelligence Act will impact product safety and risk for digital products. These laws extend liability beyond hardware to include software updates, data quality, and cybersecurity. Companies in sectors like defense and secure communications must ensure their AI products are reliable and maintain control. They also need to manage risks from complex supply chains and third-party software.

Europe's Top Selling Products in 2025 Revealed

SRP reviewed the best-selling products across European markets in 2025. In Italy, BPER Banca's Protection Certificate Memory Coupon was the highest-selling product, raising 490 million euros. This five-year capital protection certificate provides access to the Eurostoxx Utilities Index. It offers a conditional memory coupon of 1.80% every six months if the index stays at or above its starting level.

AI Fake Image Spreads Bondi Shooting False Flag Claim

BBC Verify is investigating an AI-generated image used to spread false claims that the Bondi Beach shooting in Sydney, Australia, was staged. The fake picture shows an injured man, resembling victim Arsen Ostrovsky, with fake blood. This image has been shared millions of times on social media to promote a "false flag" conspiracy. BBC Verify confirmed the image is AI-generated by noting scrambled text on a shirt and inconsistent bloodstains compared to real photos. Separately, new dashcam footage of the attack was verified and showed a man confronting an attacker, Sajid Akram, and a flag linked to the Islamic State group.

OpenAI Boosts Hiring for Future Acquisitions

OpenAI recently hired Kim Lee, a top corporate development executive from Google, suggesting the company might be planning more acquisitions. Lee will lead corporate development, aiming to move quickly on deals. This follows other major business-side recruitments this month. OpenAI has already made at least five major acquisitions in 2025. The company also finalized its recapitalization process in October, potentially setting the stage for an initial public offering in the new year.

Fortinet Adds Security to Nvidia AI Factories

Fortinet now offers its FortiGate-VM security to Nvidia BlueField-3 for AI factories. This integration embeds networking and security directly into the AI factory infrastructure, allowing core security functions to run on the BlueField-3 DPU instead of the host CPU. AI factories are data centers built to develop and train AI models at scale, requiring secure, high-performance environments. This solution helps secure private cloud and AI environments without slowing down critical workloads. Kevin Deierling from Nvidia stated this extends infrastructure-offloaded services to include advanced security, ensuring secure and scalable AI-powered data centers.

Perplexity AI Updates iPad App for Research

Perplexity AI Inc. relaunched its iPad app with new features focusing on research tools and improved multitasking. The updated app, available Tuesday, targets business customers, students, and professionals by offering more citations than a regular search. This redesign responds to increased usage from business users and aims to boost subscription revenue, as advanced research features are limited to paid tiers. The app also uses the iPad's split-view, allowing users to run Perplexity alongside other apps, and features a larger sidebar for easier navigation.

Starling AIX Helps Businesses Build AI Memory

Starling AIX, developed by Chris Kincade, helps organizations create an "Organizational Brain" by blending human insight with machine capabilities. Its platform structures company knowledge so large language models can use it contextually. The Universal Cognitive Architecture provides a way to organize data, decisions, and values for consistent AI reference. The Starling System Library allows teams to build structured data sets through natural conversation. Kincade emphasizes that the goal is not to automate decisions but to ensure they are remembered, referenced, and improved upon, preserving institutional memory in the age of AI.

Alger Expert Shares 2026 AI Market Predictions

Ankur Crawford, an executive vice president and portfolio manager at Alger, shared her outlook for AI in 2026. She believes the overall economic situation is acceptable but not especially strong. Crawford expects consumers to show more financial strength during the first half of 2026. She discussed these predictions with Romaine Bostick and Katie Greifeld on "The Close."

Lambda Offers Multi-Cloud AI Infrastructure Plan

Lambda provides a multi-cloud plan for high-performance AI infrastructure, helping AI and ML teams manage growing workloads. This blueprint gives access to the latest GPU technology and offers customized solutions for better technical and financial efficiency. Using multiple clouds helps reduce risks like GPU capacity shortages, data residency issues, and high interconnect costs. Lambda offers dedicated NVIDIA GPU clusters, managed Kubernetes, and S3-compatible storage with no data-transfer fees. This allows organizations to deploy AI training and inference where policies and latency require, rather than being limited by single-cloud constraints.

Sources

• From Oracle to Broadcom, the Concerns About Artificial Intelligence Stocks Are Starting to Pile Up
• From Oracle to Broadcom, the concerns about artificial intelligence stocks are starting to pile up
• From Oracle to Broadcom, the Concerns About Artificial Intelligence Stocks Are Starting to Pile Up
• Introducing Cisco’s Integrated AI Security and Safety Framework
• Navigating product liability in high-security sectors: Addressing AI-driven risks under German and European law
• Best-selling products Europe 2025: utilities, AI and fixed income
• BBC Verify: The AI fake being used to spread 'false flag' claim about Bondi shooting
• Is OpenAI Entering Its Acquisition Era?
• Fortinet brings FortiGate-VM security to Nvidia BlueField-3 for AI factories
• Perplexity AI Revamps iPad App to Deepen Focus on Research Tools
• Starling AIX Is Redefining How Organizations Build Institutional Memory in the Age of AI
• Alger's Ankur Crawford Shares Her 2026 AI Outlook
• Lambda’s multi-cloud blueprint for high-performance AI infrastructure