Ken Griffin Warns of AI Bubble as MSCI Revenue Hits $892M

Key Takeaways

• <ul><li>Ken Griffin says investors are almost certainly in an AI bubble but remains optimistic about long-term innovation</li><li>Corporate America struggles to extract real value from AI models and is reviewing large bills from OpenAI and Anthropic</li><li>AI infrastructure investment is expected to reach $10.3 trillion by 2032</li><li>MSCI reported Q3 2026 revenue of $892 million, up 12.4% year-over-year, driven by AI-powered products</li><li>GEICO mapped legacy claims code in one day using Anthropic's Fleet Kit for $1,000</li><li>Vanguard cut research time from 7 days to 15 minutes using AI tools and MCP servers</li><li>Cisco reduces bank AI deployments from a year to weeks through strict security protocols</li><li>Meta updated commercial terms after backlash, removing scraping claims for AI training</li><li>Two U.S. senators launched an inquiry into Google's AI use in schools after a WSJ investigation</li><li>A survey of nearly 6,000 executives found many companies see productivity gains only within single departments</li></ul>

Ken Griffin says AI bubble is likely but long-term innovation remains strong

Citadel founder Ken Griffin said an AI investment bubble has very likely formed. He spoke at the Future of Global Markets conference in New York on October 21. Griffin remains optimistic about AI's long-term impact, comparing it to the dot-com bubble. He said companies need to show real productivity gains from AI spending.

Ken Griffin says AI stocks are in a bubble but worth holding

Ken Griffin said there is almost certainly a bubble in artificial-intelligence stocks. He spoke at a Citadel Securities conference on Tuesday. Griffin said radical economic changes are often tied to bubbles. He believes investors should stay invested because the biggest market leaders often emerge after bubbles clear.

Ken Griffin says AI bubble is almost certain but may produce lasting winners

Ken Griffin said investors are almost certainly in an AI bubble during Citadel's Future of Global Markets Conference on October 20. He said major economic shifts usually come with bubbles. He noted that companies surviving tough market cycles often lead markets for decades. He also said corporate America has struggled to get real value from AI models.

Ken Griffin says investors are in an AI bubble as tech debt grows

Billionaire investor Ken Griffin said investors are dealing with an AI bubble. He spoke at Citadel's Future of Global Markets Conference. He said corporate America has been uneven in extracting value from AI models. He noted companies are reviewing large bills from OpenAI and Anthropic while AI infrastructure investment is expected to reach $10.3 trillion by 2032.

Ken Griffin calls AI a bubble after Citadel bought AI fund assets

Ken Griffin warned that investors are almost certainly in an AI bubble. He spoke at Citadel's Future of Global Markets Conference in New York. Three months earlier, Citadel bought the public stock book of Situational Awareness, an AI-focused hedge fund. The deal paid off as the AI trade rebounded. Griffin's warning comes ahead of Anthropic's expected IPO after the November midterm elections.

Citadel's Ken Griffin says investors are almost certainly in an AI bubble

Ken Griffin from Citadel said investors are almost certainly in an AI bubble. He spoke at the Future of Global Markets Conference in New York. Griffin said Corporate America struggles to get value from AI models and is worried about bills from Anthropic and OpenAI. He compared the current AI boom to the dot-com era. He also noted that AI could spark a golden era for entrepreneurs while threatening white collar jobs.

MSCI Q3 2026 revenue jumps 12.4% on AI-powered products

MSCI reported Q3 2026 revenue growth of 12.4% year-over-year, driven by strong asset-based fees and record net new sales. Adjusted EPS rose 14% and the retention rate improved to 95%. The company has nearly $23 trillion in assets benchmarked to its indexes. CEO Henry Fernandez said there will be no margin expansion because MSCI will reinvest AI efficiency gains into growth. The firm aims to double or triple its size over the next decade.

MSCI Q3 results show AI-driven growth and diverse clients

MSCI beat Wall Street expectations with Q3 2026 revenue of $892 million and non-GAAP profit of $5.11 per share. Growth was led by Index, Analytics, and Private Capital Solutions. CEO Henry Fernandez highlighted momentum across these product lines. AI-powered products and agentic technology played a central role, including a new model creation factory. Hedge funds drove a 32% increase in Index subscription run rate. The Sustainability and Climate segment faced headwinds but the First Street acquisition is expected to help.

Cisco speeds up bank AI deployments from year to weeks

Cisco's Al Chen said his team cuts bank AI deploys from a year to weeks. Banks require strict security because they use air-gapped installs and serve millions of customers. The three key pillars are CVE exploit checks using Harbor tasks, disaster recovery with ClickHouse backups, and preventing data leakage through LLM judges. Chen said approved image pull processes and sourced GPUs shorten delivery while formal security reviews add months. Vertafore also launched Velocity Policy Intake and Policy Check AI agents.

Meta changes commercial terms on AI training after backlash

Meta updated its commercial terms after backlash from news groups and legal experts. The company removed claims that it could scrape websites to train AI models. The new policy allows business accounts to scrape websites if they use Facebook, Instagram, Threads, or WhatsApp. The update follows concerns that platform terms of service can override technical tools like robots.txt files. Independent researchers and publishing groups continue to push for regulation to check platform market power.

Researchers find new attack method for safe AI agents

Scientists studied a safety issue in a technology called Safe Multi-Agent Reinforcement Learning. They created a new attack method named MinMax Attack that targets groups of AI agents working together. This attack changes the information agents see to make them act unsafely while hiding the attack. Tests showed this method causes more safety problems than other attacks and is harder to detect.

GEICO uses AI to map legacy claims code in one day

GEICO spent $1,000 to map its old claims code in a single day using artificial intelligence. The company used Anthropic's Fleet Kit and code-modernization plugin to run a special work session. Engineers at GEICO had very little time for big projects, so they used AI agents to handle smaller tasks. The effort created new APIs and found a bug in their production system.

Companies struggle to see productivity gains from AI tools

Many businesses invest heavily in AI but do not see better productivity results. A survey of nearly 6,000 senior executives in early 2026 found this common problem. The research shows that current AI tools often only improve work inside single departments. Companies need tools that connect different functions to truly boost overall efficiency.

Senators launch inquiry into Google AI use in schools

Two U.S. senators sent a letter to Google CEO Sundar Pichai about AI risks for children. Sen. Richard Blumenthal and Sen. Josh Hawley started this inquiry after a Wall Street Journal investigation. The report found that Google AI tools are used in schools without proper oversight or parent knowledge. The senators want Google to explain how it plans to keep these tools safe for students.

Spectrite uses AI to connect product data across departments

Spectrite uses AI to help companies share product and packaging data easily. The company helps food and beverage firms manage information that is usually stuck in different systems. Their new AI assistant lets users access data without deep technical training. The tool also helps suppliers share information more easily with customers.

Vanguard AI tool cuts research from 7 days to 15 minutes

Vanguard modernized its cash ledger system, which handles 30 million calls daily, by exposing legacy code through an MCP server. Gareth Yoder explained that research time dropped from 7 days to 15 minutes using this approach. About 90% of the cash program engineers now use AI tools daily through multi-agent workflows. The team used change data capture and careful audits to migrate data while keeping consistency across systems.

Sources

• Citadel Founder Ken Griffin Says AI Investment Bubble Is Nearly Certain, Remains Bullish on Long-Term Innovation
• AI stocks are probably in a bubble — but might be worth holding onto, says Citadel’s Ken Griffin
• Ken Griffin: There Are “Almost Certainly Bubbles” in AI Right Now
• Investors ‘almost certainly’ in midst of AI bubble: Ken Griffin
• Ken Griffin Calls AI a Bubble After Citadel Bought the Dip - Micron Technology (NASDAQ:MU), Bloom Energy…
• Citadel's Ken Griffin said investors are 'almost certainly' in an AI bubble
• [MSCI Q3 2026 Earnings Call] MSCI Revenue Jumps 12.4% as AI-Powered Products Fuel Record Net New Sales; CEO Says 'No Margin Expansion' as Firm Prioritizes Growth
• MSCI Q3 Deep Dive: AI-Driven Product Expansion and Diversified Client Growth
• Cisco Vets Bank AI Deploys in Weeks, Not a Year
• Meta changes commercial terms on AI training after backlash
• Vulnerability Analysis for Safe Multi-Agent Reinforcement
• GEICO spent $1,000 to map legacy claims in a day
• The Deployment Failure Holding Back Your AI Investments
• Exclusive | Senators Express ‘Grave Concerns,’ Launch Inquiry into Google’s AI Push in Schools
• AI Connects ERP Data to Product and Packaging Specifications
• Vanguard Cut Research From 7 Days to 15 Minutes