AI & Economy News — 2026-09-26 (Sat) Evening News

Overview

While the market is becoming more cautious due to the expansion of AI-related capital investment and the sharp increase in electricity demand for data centers, the IMF has estimated the productivity improvement effect (raising labor productivity by up to 3.8%) (News On AIR). Discussions regarding changes in the employment structure and taxation of AI companies are in full swing at central banks and Congress.

Macroeconomic/growth trends

It has been pointed out that the AI infrastructure investment boom is boosting U.S. orders for capital goods (Wolf Street). Emerging market debt markets are also being discussed as potentially disrupted by AI (Seeking Alpha), and prominent investors are warning of the risks underlying the AI boom (thestreet.com). Some analysis suggests that AI will become an issue in the midterm elections (Newswise).

Productivity trends

An IMF study estimates that AI could boost labor productivity by up to 3.8% across industries (News On AIR, ET CIO). On the other hand, the disconnection of career ladders (promotional paths) has been pointed out as a new issue (The Edge Malaysia). At the United Nations General Assembly (UNGA 81), the impact of new international rules on AI on employment, data, and the digital economy was discussed (People Daily).

Trends in employment and labor market

While Georgia's governor says AI is creating jobs (WABE), there are also reports that entry-level jobs are shrinking and career paths are changing (Central Penn Business Journal). In the US Congress, a bill has been proposed (Quiver Quantitative) that would tax AI companies and use the funds to support employment measures. About 90,000 call center jobs in Arizona are threatened by AI and foreign competition, a union warns (AZ Family). The Asahi Shimbun reported on the reality of how AI is taking away jobs, but also creating new jobs, through India's "invisible workers" (Asahi Shimbun).

Trends in energy/power demand

In response to the rapid increase in demand for electricity from AI data centers, the state of its regulation is being discussed (The Regulatory Review). There are reports that cracks are appearing in Oracle's plans to build a large data center in New Mexico (WSJ). Massachusetts is also examining the environmental impact of AI infrastructure (The Boston Globe). Crusoe scraps plans for $1.25B data center with Boom turbines (TechCrunch) Data center company Nscale raises $3.36B ahead of IPO (SiliconANGLE).

Remarks by central banks, international organizations, and VIPs

- Bank of England: It was reported that the bank is working on using AI to read interest rate decisions ahead of the market and predict market reactions (This is Money). Governor Andrew Bailey himself was not mentioned in the statement. - ECB: Governing Council member Panetta reportedly called for the need to track income distribution using AI (UA.NEWS via Cyprus Mail). President Christine Lagarde's own remarks are not included in this headline. - World Bank: Reported to have held talks on AI with Pakistan's Punjab province (The Express Tribune). It is not specified as a personal statement by President Ajay Banga. - GZERO Media discusses how AI can help counter cyber attacks. - Individual statements from the IMF, OECD, Kristalina Georgieva, Mathias Cormann, Kazuo Ueda, Daron Acemoglu and Erik Brynjolfssky are not found in today's headlines.

Trends in corporate capital investment

Microsoft is focused on balancing Azure growth and capital investment risk (Yahoo Finance). AI capital investment, government bond yields, and geopolitical risks are discussed as warning signals to the market (Seeking Alpha). There is an estimate that hyperscaler AI capital investment may reach $1.2T in 2027 (Briefs Finance). According to estimates by Goldman Sachs, the break-even line for hyperscalar AI investments will be approximately 47 trillion yen in annual revenue (zaikei.co.jp). Some analysis suggests that Apple's advantage over Nvidia lasted only two months (shattered.io). Micron stock has risen more than 260% since the beginning of the year, and its sustainability will be tested when the company's financial results are released on September 30th (news.yahoo.co.jp).

Trade and industrial policy

Tariffs and AI were among the points of interest at the Trump-Xi Jinping summit (Nikkei Asia). It has been reported that Chinese AI models are gaining worldwide popularity, and Washington is on high alert (CNBC, The Tech Buzz). The US has lifted export restrictions on Claude, Fable, and Mythos models, and Anthropic is said to have passed one hurdle toward a major IPO (Stocktwits).

Integrated analysis: What is happening now

The story of the ``benefits of AI,'' as estimated by the IMF as a productivity improvement effect (up to 3.8%), and the story of the ``pains of AI'', such as the polarization of employment (risk of 90,000 call center jobs and reduction in entry-level jobs), are progressing simultaneously. This structure is supported by the huge capital investments made by hyperscalers (estimated at $1.2T in 2027), but skepticism about their profitability line (annual revenue of about 47 trillion yen, estimated by Goldman) is growing, and ``investment flaws'' such as cracks in Oracle's data center plan and the withdrawal of Crusoe are beginning to surface individually. The debate over regulations to address the rapidly increasing demand for electricity shows that we have entered a stage where the burden that this expansion of investment has on the real economy (power grid and local environment) cannot be ignored. On the central bank side, there are parallel moves to incorporate AI from an ``object of analysis'' into a ``tool for policy implementation,'' with the ECB's Panetta calling for AI to track income distribution, and the Bank of England using AI to determine interest rates. In the United States and China, there have been simultaneous reports of the relaxation of AI export regulations (Anthropic) and warnings about Chinese AI models (CNBC), and it can be seen that industrial policy is wavering between protection and openness.

Future points of interest

- Micron's September 30th financial results are expected to be a test of the sustainability of the capital investment boom based on AI demand. - The focus may be on how actual profits will change compared to Goldman's estimated "breakeven line" (annual revenue of about 47 trillion yen), and whether the rifts in individual projects such as Oracle will widen. - The U.S. Congress' bill to tax AI companies and labor union movements surrounding job losses, such as the Call Sen Rules, could accelerate policy debates over AI and employment.