Federal Reserve Chair Kevin Warsh says the Fed is modeling several possible effects of artificial intelligence, not predicting which will occur. AI could lift productivity and growth while keeping employment high; automate jobs and push unemployment up; or fail to produce expected returns, bringing lower company valuations, a market decline and slower growth. The outcomes could overlap and depend on how quickly AI spreads, which tasks it takes over and how workers are trained. Warsh says productivity gains have been reasonably strong but have not clearly reached a lasting higher level. AI could also affect inflation and future monetary policy, while the Fed faces other uncertainties including elevated inflation, the Iran war and tariffs. In the Fed’s financial stability report, 87% of respondents identify AI-related risks—including cyberattacks, disappointing earnings and complex financing—as the financial system’s leading risk.
Anthropic is pressing ahead with an IPO that could raise more than $60 billion, with one report putting its target at as much as $64.4 billion, alongside a planned $26 billion private placement. The Claude maker has $6.2 billion in debt and says it has committed to $900 billion in computing and infrastructure spending in coming years, including $70 billion over the next 12 months. Nvidia has agreed to buy an estimated $10 billion of Anthropic stock after the IPO. Advanced Micro Devices, Broadcom, Taiwan Semiconductor Manufacturing Company, Micron, Google and Microsoft are also named among investors supporting the offering. Investor doubts about AI valuations, borrowing costs, profitability, volatile markets and political opposition to data centers add uncertainty.
The IPO comes as AI funding attracts substantial capital but raises questions about returns. AI companies received 86% of U.S. venture capital in the first half of 2026, when more than $350 billion flowed to AI startups; venture funds returned 17.8%, with gains concentrated in a small number of large AI-linked companies. A separate S&P Global survey found 70% of more than 1,100 limited partners plan to increase AI investment. Yet S&P cautions that private-market performance often reflects higher valuations rather than cash returns. Funds closed in the first half of 2026 raised $312 billion, and among S&P Composite 1500 companies discussing AI on earnings calls in the first nine months of 2026, only five reported investment figures alongside actual or expected returns. S&P expects measurable results to matter more to investors in 2027.
Companies are also putting AI into products and operations. Austrian fintech Bitpanda has created a $10 million unit, led by AI and machine-learning expert Dr. Christian Fries, to connect AI assistants with its trading platform. Planned features include automated trading tasks, portfolio management and real-time market analysis, with integration expected within 12 months. Passenger-communications software provider 15below appointed Vishal Manani as chief product and technology officer and Kemal Ahmed as product director. Its systems serve more than 150 airlines, airports and rail operators and sent nearly 1.3 billion notifications in 2025; over the next three years it plans to expand reservation partnerships, add AI-powered personalisation and messaging, and develop its IROPS disruption platform.
Microsoft has added generative AI models and tools to Foundry, its platform for building, testing and deploying AI applications and agents. New improvement loops use evaluations, expert human ratings and real-world results to assess quality, safety, latency and cost. Foundry also gains an updated decision model based on Google’s Gemma 4 26B A4, with image-analysis support, alongside Microsoft’s MAI coding, image and cybersecurity models. Its expanded catalogue includes the top 50 models and more than 11,000 models overall. Separately, researchers tested an expanded Cognitive Pairwise Comparison framework that structures human preferences for AI and data-science decisions, using movie recommendation and the MovieLens dataset across supervised learning, dimensionality reduction and unsupervised learning.
AI’s economic and infrastructure effects are also playing out internationally. India’s chief economic adviser, Anantha Nageswaran, says the country could benefit if the AI investment trade unwinds: India was largely left out of the recent boom, whose foreign capital inflows strengthened the rupee and made global competition harder for Indian businesses. Across South and Southeast Asia, open-weight Chinese models such as DeepSeek and Alibaba’s Qwen are gaining users, including among governments wary of Beijing. Downloadable models alone may not create strong leverage, but dependence could grow through cloud services, computing infrastructure and surrounding tools. Alibaba opened a Singapore data centre in 2025, while India has committed US$1.2 billion to its National Supercomputing Mission using Indian hardware; it still relies on imported GPUs, mostly from U.S. chipmakers. Some adopters use both U.S. and Chinese models to limit lock-in.
In security, FF News reports that a quantum computer has broken a 43-bit encryption key, raising concerns that more powerful quantum computers could threaten stronger encryption and the security of online transactions and data, including in finance. The report points to the need for new forms of encryption.
Key Takeaways
- Kevin Warsh says the Fed is modeling AI’s potential effects—not making predictions—including productivity-led growth, job displacement and losses if investments disappoint.
- In the Fed’s financial stability report, 87% of respondents rank AI-related risks, including cyberattacks and complex financing, as the top risk to the financial system.
- Anthropic is proceeding with an IPO that could raise more than $60 billion; reports also cite a target of up to $64.4 billion and a planned $26 billion private placement.
- Anthropic reports $6.2 billion in debt and $900 billion in planned computing and infrastructure commitments, including $70 billion over the next 12 months; Nvidia plans an estimated $10 billion stock purchase after the IPO.
- Advanced Micro Devices, Google and Microsoft are among the named supporters of Anthropic’s fundraising, alongside Broadcom, TSMC and Micron.
- AI attracted 86% of U.S. venture capital in the first half of 2026, while S&P warns that private-market valuation gains have not necessarily produced cash returns; it expects investors to prioritize measurable results in 2027.
- Bitpanda is investing $10 million in AI trading tools, while 15below plans AI personalisation and messaging as it serves more than 150 airlines, airports and rail operators.
- Microsoft expands Foundry with human- and outcome-based improvement loops, new MAI models and more than 11,000 available models; its updated decision model is based on Google’s Gemma 4 26B A4.
- India’s Anantha Nageswaran says an AI investment pullback could ease pressure on Indian businesses; regional adoption of Chinese models is growing, while reliance on cloud and computing infrastructure remains a potential source of dependence.
- A reported quantum-computing demonstration broke a 43-bit encryption key, renewing concern about future encryption security.
Warsh Says Fed Is Modeling AI’s Possible Economic Effects
Federal Reserve Chairman Kevin Warsh spoke Monday at a San Francisco symposium about how artificial intelligence could affect the economy. He said Fed staff are studying possible scenarios, but stressed that they are examples, not predictions. Warsh did not discuss the near-term outlook for interest rates or the economy. Before becoming Fed chairman this year, he invested in early-stage technology companies, and he has said AI-driven productivity gains will be important to the U.S. economy.
Fed Analysts Explore Three Possible Paths for AI and the Economy
A report describes three possible ways artificial intelligence could affect the economy. AI could boost productivity and growth, replace workers and raise unemployment, or fail to deliver expected gains and slow economic growth. The report says these outcomes could overlap and will depend on how quickly AI spreads, which tasks it takes over, and how workers are trained. It says the effects remain uncertain and that more research is needed.
Warsh Warns AI Could Bring Jobs Losses or a Market Drop
Federal Reserve Chairman Kevin Warsh discussed AI’s possible effects on the economy at a San Francisco Fed conference. He said AI could raise productivity and growth, but it could also cause high unemployment if technology changes the labor market. If AI fails to deliver expected productivity gains, high company valuations could fall and a stock market decline could slow the economy. Warsh said the Fed is studying different scenarios because AI may also affect future monetary policy, though its effects and timing remain unclear.
Fed Models AI’s Economic Risks as Warsh Calls for Planning
Federal Reserve Chairman Kevin Warsh said Monday that the Fed is modeling how artificial intelligence could affect the economy, from boosting productivity to replacing workers or triggering a market correction. AI already accounts for a significant share of growth, investment and stock market gains, but its longer-term effects remain uncertain. The Fed’s semiannual financial stability report said 87% of respondents viewed AI risks as the top risk to the financial system, including cyberattacks, disappointing earnings and complex AI-related financing. Warsh said AI-driven productivity could help slow inflation, but recent gains have not clearly risen to a higher long-term level. He said the Fed should prepare for different outcomes without supporting or opposing AI.
Warsh Says Fed Must Prepare for AI’s Uncertain Economic Impact
Federal Reserve Chairman Kevin Warsh said Monday that the Fed is modeling how artificial intelligence could reshape the U.S. economy. AI could boost productivity and growth, but it could also displace many workers or lead to a market correction if investments disappoint. The Fed’s semiannual financial stability report found that 87% of respondents named AI-related risks, including cyberattacks and complex financing, as the top risk to the financial system. Warsh said productivity gains have been reasonably strong but have not clearly accelerated in a lasting way, and he called for planning for possible outcomes.
Fed prepares for AI’s best and worst economic outcomes
Federal Reserve official Kevin Warsh said the Fed is planning for several possible effects of artificial intelligence. AI could raise productivity and share economic gains among workers and business owners, or it could replace jobs and increase unemployment. A third possibility is that AI investments fail to deliver expected returns, causing an economic slowdown and job losses. Warsh said the Fed cannot predict the future, but must prepare because AI could affect the economy and monetary policy.
Fed models AI’s possible effects on the US economy
Federal Reserve Chair Kevin Warsh said the central bank is studying how artificial intelligence could reshape the U.S. economy. The possible outcomes include higher productivity and growth, job losses from automation, or disappointing investment returns that lead to a market decline. The Fed is weighing these risks as it faces elevated inflation and other uncertainties, including the war with Iran and tariffs. Warsh said recent productivity gains have not yet shown a clear, lasting rise, and that the Fed’s job is to prepare for different outcomes.
Warsh warns AI could boost growth or disappoint
Federal Reserve Chair Kevin Warsh outlined three possible paths for artificial intelligence at a meeting of the Fed’s AI task force at the Federal Reserve Bank of San Francisco. AI could raise productivity and economic gains while keeping the U.S. close to full employment, or it could replace workers and push unemployment higher. Warsh also said AI companies could fail to earn enough from their investments, leading to falling stock values, an economic slowdown, and more unemployment. He said he was not making predictions, but stressed the importance of American leadership in AI.
Anthropic presses ahead with IPO as AI market worries grow
Anthropic plans to move forward with an initial public offering that could raise more than $60 billion on Thursday, despite nervous investors, rising financing costs, and political opposition to data centers. The Claude chatbot maker has $6.2 billion in debt and is unprofitable under standard accounting rules. It says it has commitments to spend $900 billion on computing and infrastructure in coming years, including $70 billion in the next 12 months. Nvidia has agreed to buy an estimated $10 billion of Anthropic stock after the IPO, while the article notes that volatile AI stocks and a history of IPO delays add uncertainty.
Anthropic moves ahead with IPO amid investor doubts
Anthropic is proceeding with an initial public offering that could raise more than $60 billion on Thursday, even as investors worry about an AI bubble and the costs of funding the industry rise. The company makes the Claude chatbot and has taken on $6.2 billion in debt, which it describes in financial filings as substantial. AI stock volatility could make investors wary, while higher borrowing costs and concerns about profitability add pressure. Nvidia has agreed to buy an estimated $10 billion of Anthropic stock after the IPO.
Anthropic pushes ahead with a huge IPO amid market fears
Anthropic plans to move forward with its initial public offering despite worries about an AI bubble, shaky markets and rising borrowing costs. The Claude chatbot maker could raise more than $60 billion, while Nvidia has agreed to buy an estimated $10 billion of its stock after the IPO. Anthropic has $6.2 billion in debt and says it has signed deals requiring $900 billion in computing and infrastructure spending in coming years, including $70 billion over the next 12 months. Political opposition to new data centers adds uncertainty, but the company may want to sell shares before a possible crackdown begins.
Anthropic IPO could test confidence in AI startups and venture funds
Anthropic’s planned IPO could shape how investors view AI startups and the funds that have backed them. AI companies received 86% of U.S. venture capital in the first half of 2026, and more than $350 billion flowed into AI startups. Anthropic is seeking as much as $64.4 billion in an IPO and $26 billion in a private placement, with support from investors including Advanced Micro Devices, Broadcom, Nvidia, Taiwan Semiconductor Manufacturing Company, Micron, Google and Microsoft. Venture funds returned 17.8% in the first half of 2026, but gains have been concentrated in a small number of large AI-linked companies.
AI investment draws private funds, but cash returns remain uncertain
A survey by S&P Global Market Intelligence found that 70% of more than 1,100 limited partners plan to increase their AI investments. But S&P warned that much of AI’s apparent private-market performance comes from higher valuations that have not yet produced cash returns. Funds that closed in the first half of 2026 raised $312 billion, while investors have raised concerns about concentrated funding, crowded investments and data center risks. Among S&P Composite 1500 companies, nearly 80% discussed AI on earnings calls in the first nine months of 2026, but only five reported investment figures alongside actual or expected returns. S&P expects investors in 2027 to favor measurable results over AI exposure alone.
Bitpanda creates a $10 million unit to add AI tools to trading
Austrian fintech company Bitpanda has created a $10 million division to develop and add AI technology to its trading platform. Dr. Christian Fries, an expert in AI and machine learning, will lead the division. The tools are intended to let users connect AI assistants to the trading interface, automate trading tasks, manage portfolios and receive real-time market analysis. Bitpanda CEO Eric Demuth said the goal is to make trading easier, more efficient and more accessible. The integration is expected to be completed within the next 12 months.
India may gain as AI investment trade unwinds, says Anantha Nageswaran
India’s chief economic advisor, Anantha Nageswaran, said the country could benefit as the AI investment trade unwinds. He said India was largely left out of the AI investment boom of the past two years and that the boom brought foreign capital that raised the rupee’s value and made it harder for Indian businesses to compete globally. Nageswaran called the unwinding a possible relief that could let India focus on its own economic growth and development. His comments may inform debate on the country’s economic policy.
15below strengthens product leadership as it expands AI investment
15below appointed Vishal Manani as Chief Product and Technology Officer and hired Kemal Ahmed as Product Director. Manani will lead product, engineering, AI, client delivery and technology operations at the Brighton-based passenger communication software company. Ahmed brings more than 15 years of travel technology experience, including roles at NATS and SITA. The company’s systems serve more than 150 airlines, airports and rail operators and sent nearly 1.3 billion notifications in 2025. Over the next three years, 15below plans to expand reservation system partnerships, add AI for personalisation and messaging, and develop its IROPS disruption platform.
A human guided framework improves AI and data science comparisons
Researchers proposed an expanded Cognitive Pairwise Comparison framework to help people express preferences and priorities in artificial intelligence and data science work. Human input can add expert knowledge and help address limits of fully automated machine learning, but tasks such as choosing features and models can take many rounds of work. The framework aims to measure user preferences in a structured way. The study tested its use with movie recommender systems and the MovieLens dataset, covering supervised learning, dimensionality reduction and unsupervised learning.
Quantum computer breaks a 43 bit encryption key
FF News reports that a quantum computer has been used to break a 43 bit encryption key, marking a milestone for quantum computing and AI security. The article says stronger quantum computers could break stronger encryption keys. This could affect the security of online transactions and data, including in the financial industry. The report also points to the need for new forms of encryption.
Microsoft adds AI models and tools for ongoing agent improvement
Microsoft announced new generative AI models and updates to Foundry, its platform for building, testing and deploying AI applications and agents. Foundry now supports improvement loops that use evaluations, human expert ratings and real world results to measure quality, safety, latency and cost. Microsoft also introduced an updated decision model based on Google’s Gemma 4 26B A4, with image analysis support. Its new MAI models include coding, image and cybersecurity tools, while Foundry’s expanded model pool includes the top 50 models and more than 11,000 models overall.
Chinese AI models are spreading in Asia but dependence is not certain
Chinese open weight models such as DeepSeek and Alibaba’s Qwen are gaining users across South and Southeast Asia, including governments concerned about Beijing. The models can be downloaded and changed, so they do not by themselves give China strong leverage, but reliance could grow through cloud services, computing infrastructure and the tools built around a model. Alibaba opened a data centre in Singapore in 2025, while India is investing US$1.2 billion in a National Supercomputing Mission using Indian hardware. The article says India still depends on imported GPUs, mostly from US chipmakers. It also notes that US models compete for users and that some adopters use both US and Chinese models to reduce the risk of lock in.
Sources
- Warsh Says Fed Is Gaming Out How AI Could Reshape Economy
- Federal Reserve Chair Outlines Three AI Economic Scenarios
- FED Chairman Kevin Warsh Issues Warning on Artificial Intelligence: “A Major Economic Crisis Could Occur”
- AI could transform the economy, but the Fed is struggling to predict how
- AI could transform the economy, but the Fed is struggling to predict how
- Fed Builds Plans for Best and Worst AI-Driven Economic Outcomes
- AI could transform the economy, but the Fed is struggling to predict how
- Warsh says AI could ‘disappoint’
- Anthropic forges ahead with its IPO even as AI fears mount and markets wobble
- Anthropic forges ahead with its IPO even as AI fears mount and markets wobble
- Anthropic forges ahead with its IPO even as AI fears mount and markets wobble
- Anthropic IPO: For hundreds of billions of dollars flowing into AI startups, a big debut is a must
- Private Market Investors Aren't Seeing a Return On AI Capital
- Dealroom.co | Bitpanda plugs AI agents into trading via $10M division
- India to benefit from unwinding of AI trade, says CEA Anantha Nageswaran
- 15below appoints product leaders amid AI investment
- Advancing Cognitive Pairwise Comparisons for Human-Centric Artificial Intelligence and Data Science: A Human-in-the-Loop Framework
- FF News
- Microsoft rolls out new first-party models and updates to agentic AI continuous improvement
- Chinese AI dominance is not Asia’s destiny
Comments
Please log in to post a comment.