Private market appetite for AI-related investments is hitting record levels. A Dynamo Software survey found 63 percent of limited partners plan to increase alternative investment allocations, while a Morningstar study showed 71 percent of asset owners consider generative AI a material investment issue. AI valuations and capital spending concerns ranked as the top macro-market worry at 73 percent, and North America is regaining ground as the leading destination for alternative capital.
Nvidia CEO Jensen Huang forecast that chip sales will roughly double over the next year as AI demand spreads across industries. He made the remarks at a summit with U.K. King Charles III in Scotland. The projection upgrades last month's estimate of 70 percent revenue growth, and Nvidia shares rose as much as 2.8 percent on the news. Nvidia controls about 80 percent of the AI accelerator market, though supply chain capacity remains the main constraint.
Big Tech companies are reaping substantial tax breaks for AI spending. Amazon, Alphabet, Meta, Microsoft, and Oracle together received $70 billion in tax breaks last year. A 2025 law called the One Big Beautiful Bill Act allows companies to write off equipment costs quickly, with $27 billion going toward accelerated depreciation. Critics argue these firms do not need government subsidies, especially given public polling showing Americans distrust AI and data centers.
On the cybersecurity front, Atos partnered with GCH to open a new AI-driven security operations center in the UAE. The facility will provide 24/7 monitoring, threat intelligence, and incident response for critical infrastructure and is expected to be operational by year-end. Separately, the U.K. government released a free AI risk management toolkit, available online and in print, designed to help organizations of all sizes identify and reduce AI-related risks through a structured assessment framework.
Retirement advisors are moving past the experimental phase with AI. T. Rowe Price's annual Defined Contribution Consultant Study found the share of firms saying AI use is too early to evaluate fell from 44 percent in 2025 to 14 percent in 2026. Seventy-eight percent now use AI for internal workflows and 67 percent for client meeting prep. Meanwhile, KnowBe4 researchers demonstrated that hidden instructions embedded in an email can hijack AI assistants using indirect prompt injection. In a test involving Google Workspace and Gemini, an AI workflow extracted a one-time passcode from an inbox without the user opening the malicious email, though later Gemini models resisted the same attack.
Key Takeaways
- <ul><li>63 percent of limited partners plan to increase alternative investment allocations, the strongest reading in Dynamo Software's survey history</li><li>Nvidia CEO Jensen Huang forecasts chip sales will roughly double next year, upgrading last month's 70 percent revenue growth projection</li><li>Amazon, Alphabet, Meta, Microsoft, and Oracle received $70 billion in tax breaks tied to AI equipment spending under the One Big Beautiful Bill Act</li><li>Atos and GCH opened an AI-driven security operations center in the UAE offering 24/7 monitoring for critical infrastructure</li><li>The U.K. government released a free AI risk management toolkit with a risk assessment framework for organizations of all sizes</li><li>T. Rowe Price's study shows 78 percent of retirement advisors now use AI for internal workflows, up from earlier experimental hesitancy</li><li>Nvidia controls about 80 percent of the AI accelerator market, with supply chain capacity as the main constraint on growth</li><li>Hidden email instructions can trick AI assistants into leaking sensitive data through indirect prompt injection, as demonstrated in a Google Workspace and Gemini test</li><li>73 percent of asset owners rank AI valuations and capital spending as their top macro-market concern, per Morningstar</li><li>Later Gemini models showed improved safeguards against the prompt injection attack that earlier versions failed to resist</li></ul>
Private market appetite hits record high as AI risk reshapes portfolios
Two new surveys show private market allocations are rising to record levels. Dynamo Software found 63 percent of limited partners plan to increase alternative investment allocations, the strongest reading in its history. Morningstar found 71 percent of asset owners see generative AI as a material investment issue. AI valuations and capital spending concerns topped macro-market worries at 73 percent. North America is regaining ground as the top destination for alternative capital.
DC consultants shift from testing AI to putting it to work
T. Rowe Price's sixth annual Defined Contribution Consultant Study shows retirement advisors are now using AI in daily operations. The share of firms saying AI use is too early to evaluate fell from 44 percent in 2025 to 14 percent in 2026. Seventy-eight percent use AI for internal workflows and 67 percent for client meeting prep. Private credit interest rose sharply, with a likelihood score of 2.6 in 2026 up from 1.7 in 2024. Cryptocurrency is mostly seen as suited for self-directed brokerage windows.
Jensen Huang says Nvidia will sell twice as many chips next year
Nvidia CEO Jensen Huang said the company expects to sell twice as many chips next year as it does this year. He made the comments at a summit with U.K. King Charles III in Scotland. Huang said AI is driving demand across industries and countries. Nvidia makes GPUs, CPUs, switch chips, and other products. The forecast points to continued massive growth for the next six quarters.
Nvidia's Huang Sees Chip Sales Doubling as AI Demand Outpaces Supply
Nvidia CEO Jensen Huang said chip sales will roughly double over the next year as AI demand spreads across industries. The forecast upgrades last month's projection of 70 percent revenue growth. Nvidia shares rose as much as 2.8 percent on the news. Nvidia controls about 80 percent of the AI accelerator market. Supply chain capacity is the main constraint, with manufacturing partner TSMC being a key factor.
Atos Partners With GCH To Open New AI-Driven Security Operations Center In UAE
Atos partnered with GCH to open a new AI-driven security operations center in the UAE. The center will provide advanced security solutions for critical infrastructure. It will offer 24/7 monitoring, threat intelligence, and incident response services. The center is expected to be operational by the end of the year. The partnership aims to strengthen cybersecurity capabilities in the Middle East.
U.K. launches AI risk management toolkit for organizations
The U.K. government released a new AI risk management toolkit to help organizations identify and reduce risks from artificial intelligence. It includes a risk assessment framework and guidelines developed with industry experts. The toolkit is free, available online and in print, and works for organizations of all sizes. It supports the government's goals of transparency, accountability, and responsible AI use.
Hidden email instructions can hijack AI assistants, researchers show
KnowBe4 researchers found that hidden instructions in an email can trick an AI assistant into leaking sensitive information. In a test using Google Workspace and Gemini, an AI workflow extracted a one-time passcode from an inbox without the user reading the malicious email. The attack used indirect prompt injection, where malicious content was embedded in an email's HTML. Later Gemini models resisted the same injection, showing improved safeguards.
Big Tech companies reap huge tax breaks for AI spending
Five major tech companies, including Amazon, Alphabet, Meta, Microsoft, and Oracle, received $70 billion in tax breaks last year. A 2025 law called the One Big Beautiful Bill Act allows companies to write off equipment costs very quickly. Of the tax breaks, $27 billion went toward accelerated depreciation. Critics argue that these companies do not need government subsidies for AI investment, especially since public polling shows Americans distrust AI and data centers.
Tempus AI stock sees strong trading volume amid buy questions
Tempus AI, listed on NASDAQ under the ticker TEM, is experiencing high trading volume. The company provides AI solutions for healthcare, finance, and education. Its stock has been performing well in recent months. Experts are analyzing market trends to determine whether the stock represents a good investment opportunity.
AI infrastructure demand stays strong despite slowdown calls
A tech fund manager says demand for AI infrastructure will keep growing even if OpenAI and Anthropic slow their progress. Second- and third-tier AI startups will continue to drive spending on infrastructure. Big tech companies are still investing heavily in AI resources. The fund manager believes the slowdown of two major AI labs will not significantly reduce overall infrastructure demand.
Building Trust in AI for Railway Safety Applications
AI is currently limited to non-safety critical uses in railways because of strict industry standards. Researchers propose focusing on three areas to increase trust in AI: robustness, Operational Design Domain, and explainability. Robustness ensures AI systems perform reliably under all conditions per ISO24029. ODD defines operating conditions per DIN DKE SPEC 99004. Explainability helps humans understand AI decisions. The authors argue that meeting these trust requirements will unlock AI's full potential in safety-critical railway and other mission-critical domains.
Sources
- Private market appetite hits record high as AI risk reshapes portfolios
- DC consultants shift from testing AI to putting it to work
- Jensen Huang says Nvidia will sell twice as many chips next year
- Nvidia's Huang Sees Chip Sales Doubling as AI Demand Outpaces Supply
- Atos Partners With GCH To Open New AI-Driven Security Operations Center In UAE
- U.K. AI risk management toolkit targets current and future compliance
- Could an Email You Never Read Hijack Your AI Assistant?
- Big Tech Is Raking in Enormous Tax Breaks for AI Development
- Tempus AI (NASDAQ:TEM) Sees Strong Trading Volume
- OpenAI, Anthropic Slowdown Call Won't Derail AI Infrastructure Demand, Why One Fund Manager Sees No Spending Pause in Sight
- Building Trust in Artificial Intelligence: A Necessity for Railway Applications
Comments
Please log in to post a comment.