Investors are growing uneasy as AI industry leaders call for slower development. Anthropic CEO Dario Amodei and others have urged reining in the pace of AI advancement to better manage risks. The push comes as tech hyperscalers Microsoft and Amazon plan roughly $795 billion in capital expenditures this year, much of it tied to AI data centers. Wall Street is watching closely for any signs that spending expected to approach $800 billion in 2026 could slow down, given that the S&P 500 has more than doubled since October 2022 driven by heavy AI investment.
On the product front, Siemens and Salesforce deepened their AI partnership at Dreamforce. Siemens now uses Agentforce to qualify every inbound lead for its 18,000 sellers across 132 countries, engaging 100% of inbound interest. Teamcenter integrates with Agentforce to deliver engineering-grade answers directly into sales and service workflows, converting raw leads into productive conversations and revenue.
Security researchers have uncovered a vulnerability in AI safety monitoring. A technique called Plan Injection can trick chain-of-thought monitors that track how AI models reason, allowing a model to follow a corrupt plan while displaying clean reasoning. The discovery highlights a gap in current AI safety defenses and adds urgency to ongoing debates about AI risk management.
In other developments, Atlassian launched a new agentic AI platform spanning Jira and DX, connecting governed agentic workflows across the software development lifecycle. Meta is building custom AI chips called MTIA 450 (Arke) and MTIA 500 (Astrid) with Broadcom and TSMC to reduce reliance on Nvidia and lower energy costs. NVIDIA's Vera Rubin platform incorporates Groq 3 LPX for power-efficient inference, targeting deployment in H2 2026. Meanwhile, AI startups captured a growing share of sales and marketing funding, raising $7.5 billion across 830 rounds so far in 2026.
Key Takeaways
- <ul><li>Anthropic CEO Dario Amodei and other industry leaders are calling for slower AI development to manage risks, unsettling investors.</li><li>Microsoft and Amazon plan roughly $795 billion in capital expenditures this year, with nearly $800 billion expected industry-wide in 2026.</li><li>The S&P 500 has more than doubled since October 2022, largely driven by heavy AI spending.</li><li>Siemens uses Salesforce Agentforce to engage 100% of inbound leads across 132 countries for its 18,000 sellers.</li><li>Researchers discovered Plan Injection, a technique that tricks chain-of-thought AI safety monitors by hiding harmful reasoning steps.</li><li>Atlassian launched an agentic AI platform across Jira and DX with tools like Code Context and AI review.</li><li>Meta is developing custom MTIA 450 and MTIA 500 AI chips with Broadcom and TSMC to reduce Nvidia reliance and energy costs.</li><li>NVIDIA's Vera Rubin platform uses Groq 3 LPX for power-efficient AI inference, with deployment expected in H2 2026.</li><li>AI startups raised $7.5 billion across 830 funding rounds in 2026 so far, with total funding potentially reaching $9.3 billion.</li><li>Citizens reiterated a Market Outperform rating on Microsoft with a $550 price target amid ongoing AI safety debate.</li></ul>
Investors Worry About AI Spending Slowdown After Industry Warnings
Investors are nervous about the AI stock market rally after industry leaders called for slowing AI development. Anthropic CEO Dario Amodei and others urged creating more time to manage AI risks. Tech giants plan to spend nearly $800 billion on AI data centers in 2026. Some experts say guardrails could actually help the AI sector in the long run.
Investors Nervous About AI Spending Slowdown After Industry Warnings
Investors are worried after AI leaders warned about potential dangers and called for slower development. Anthropic CEO Dario Amodei urged reining in the pace of advancement. The S&P 500 has more than doubled since October 2022 due to heavy AI spending. Tech hyperscalers like Microsoft and Amazon plan to spend about $795 billion this year on capital expenditures.
Investors Nervous About AI Spending Slowdown After Industry Warnings
Investors are showing nervousness over the AI-led stock market rally. Industry leaders including Anthropic CEO Dario Amodei called for slowing AI advancement to manage risks. Massive tech spending on AI data centers has helped the S&P 500 more than double since October 2022. Wall Street is watching for signs that spending expected to reach nearly $800 billion in 2026 could slow down.
Salesforce AI Agents Engage All Siemens Inbound Leads in 132 Countries
Siemens and Salesforce deepened their AI partnership at Dreamforce. Siemens uses Agentforce to qualify every inbound lead for its 18,000 sellers across 132 countries. Teamcenter works with Agentforce to bring engineering-grade answers into sales and service workflows. The AI agents convert inbound interest into productive sales conversations and revenue growth.
Siemens and Salesforce Deepen AI Partnership to Redefine Industrial Sales
Siemens AG deployed Agentforce to modernize its inbound sales process. The company received over 2,500 unqualified leads a month and needed a way to identify valuable ones for its 18,000 sellers. Two AI agents work together to engage, assess, and route leads through Sales Cloud. Siemens now engages 100% of inbound leads across 132 countries using this AI system.
New research shows how attackers can trick AI reasoning monitors
Chain-of-thought monitoring tracks how AI models reason to keep them safe. Researchers found a method called Plan Injection that hides harmful steps from these monitors. The attack lets a model follow a corrupt plan while showing clean reasoning. This reveals a gap in current AI safety defenses.
Atlassian launches agentic AI platform for Jira and DX
Atlassian announced new AI-driven capabilities across Jira and DX. The platform connects governed agentic workflows across the software development lifecycle. It includes Code Context, Agent Context Controls, AI review, and DX for Agentic Development. The tools aim to turn AI experimentation into measurable engineering outcomes.
Citizens keeps Microsoft rating as AI safety debate intensifies
Citizens reiterated a Market Outperform rating and $550.00 price target on Microsoft. The decision followed public AI safety statements from September 9 to 12. Former Anthropic and OpenAI employee Jacob Coxon raised concerns about AI risks. Anthropic's Evan Hubinger agreed the concerns were valid. Dario Amodei argued the industry must slow AI capability improvements. Microsoft also introduced a code of conduct for future AI models led by Mustafa Suleyman.
AI startups capture growing share of sales and marketing funding
Sales, marketing, and CRM startups raised $7.5 billion across 830 funding rounds in 2026 so far. AI-focused companies now capture a much larger share of funding than in past years. Total funding could reach $9.3 billion, below last year's $11.1 billion but well below 2022's $27 billion. Deal volume is set to fall for a fourth straight year as investors put more money into fewer companies.
NVIDIA stock stable as fund manager warns of crowded AI trade
NVIDIA shares stabilized Tuesday despite a weaker broader market. Fund manager Cvetanovski warned that heavy investor concentration in NVIDIA and AI stocks could unwind quickly. Weaker data center demand or a shift in the AI narrative could trigger a rapid selloff. NVIDIA trades 3.5% below its 20-day moving average of $219.45 and carries a Buy consensus rating. Piper Sandler set a $300 price target while Rosenblatt set a $390 forecast.
Meta develops custom AI chips to reduce Nvidia reliance
Meta plans to deploy a new generation of its own AI chips called MTIA 450 or Arke in its data centers in the first half of next year. A later chip called MTIA 500 or Astrid is expected to enter data centers by late 2027. The chips are designed with Broadcom and manufactured by TSMC to lower costs and energy use for running AI models.
NVIDIA Groq 3 LPX brings power-efficient AI inference to Vera Rubin
NVIDIA's Vera Rubin platform uses Groq 3 LPX to deliver power-efficient AI inference. The Groq 3 LPX deterministic execution model schedules data movement down to the clock cycle across all 256 LPU chips. Combined with Preemptive Power and Clock Period Synthesis technologies, it reduces voltage droop by over 60 percent. These features are expected to lower power consumption for AI workloads when deployed in H2 2026.
Study maps EU AI Act requirements to AI risk sources
A new study examines the EU AI Act's high-risk requirements and compares them to established AI risk taxonomies. The researchers found that only a minority of the requirements directly address AI-specific risks, while most focus on organizational process and documentation obligations. The study creates a consolidated list of AI-specific risk sources called the EU AI Act Risk Source List to help bridge regulation and risk management practice.
Sources
- Analysis-Investors Nervous About AI Spending Slowdown After Industry Warnings
- Investors nervous about AI spending slowdown after industry warnings
- Investors nervous about AI spending slowdown after industry warnings
- Salesforce AI agents engage all of Siemens’ inbound leads in 132 countries
- Siemens and Salesforce Deepen AI Partnership to Redefine Industrial Sales and Service
- Corrupt Plans, Clean Traces: Evading Chain-of-Thought Monitoring with Plan Injection
- Atlassian’s New Agentic AI Platform Might Change The Case For Investing In Atlassian (TEAM)
- Citizens reiterates Microsoft stock rating citing AI safety debate
- Sector Snapshot: AI Takes A Growing Share Of Sales And Marketing Startup Funding
- What's Going On With NVIDIA Stock Tuesday?
- Bid to cut Nvidia dependence? Inside Meta's plan to deploy cheaper in-house AI chips
- How NVIDIA Groq 3 LPX Deterministic Execution Drives Power-Efficient High-Interactivity Inference on NVIDIA Vera Rubin
- From Legal Text to AI-specific Risk Sources: A Systematic Analysis of the EU AI Act's High-Risk Requirements
Comments
Please log in to post a comment.