AI & Finance News — 2026-09-25 (Fri) Morning News

Overview

As AI agents enter the implementation stage in payments, transactions, and insurance underwriting, banks are acting on both security concerns and legacy renewal. In the insurance field, increased costs and appraisal conflicts due to AI have surfaced, and BlackRock is expanding investment in AI infrastructure.

Bank trends

Banks are warning that AI shopping agents increase the risk of fraud, abuse, and data leaks (Fox Business). It was reported that fintech companies tend to be preferred over traditional banks due to AI models (American Banker). According to an analysis (PYMNTS.com), AI is expanding the options for upgrading banks' core systems. Cyera, a company that helps secure AI agents, raises $400m from Goldman Sachs (FinTech Global).

Insurance trends

The NYTimes reported that an appraisal battle between hospital AI and insurance company AI is pushing up medical costs. Vox points out that AI is already involved in determining whether or not to receive medical benefits. Reuters reports that AI tools have resulted in approximately $1 billion in additional costs for Blue Cross insurers. Misinformation and deepfakes are the most commonly reported AI coverage risks for insurance companies, Risk & Insurance reported. InsuranceNewsNet also reported concerns that AI could lower employee morale.

Trends in asset management and market transactions

It was reported that BlackRock is evolving its AI-related investment stance by reorganizing its model portfolio (Bloomberg). Yahoo Finance reports that BlackRock has invested $25 billion in AI data centers in Asia. BlackRock has expressed the view that stablecoins can become an AI payment layer that connects commerce and computational resources (AI Insider). Wealth Management has introduced a new AI agent. On the trading side, Public is linking AI trading agents with prediction market Kalshi (Fortune, Barron's), and FirstMark and Chemistry have invested in a startup building a dedicated exchange for AI computational resources (WSJ). Bloomberg analyzes that the IPO of new cloud company Nscale shows that it is more than an AI intermediary.

Fintech/payment trends

Stripe's AI executive introduced the growth of the AI company, with a growth rate of 175% and 48% of sales outside of its home country (saastr.com). Cleverbridge announced that it has implemented the first AI agent-driven payment in France with Visa and Revolt (Business Wire). Mastercard is working to connect SME growth, AI, and cybersecurity (FinTech Magazine). Kredete acquires agent stablecoin infrastructure Gravv (FinTech Futures).

Remarks from key figures and authorities in the financial world

No specific statements or actions by the key figures tracked in this report (Jamie Dimon, Larry Fink, Jane Fraser, David Solomon, Brian Moynihan, Ted Pick, Ken Griffin, and Paul Atkins) were identified in the headlines. There have been reports about BlackRock (model portfolio restructuring, $25 billion investment in Asia AI data center, stablecoin positioning), but Larry Fink himself has not mentioned anything. It was reported that Franklin Templeton's CEO gave a "seven-word opinion" to AI stock investors, but details of his name and title and the content of his remarks were not listed in the headline (thestreet.com).

Trends in AI-related stocks and funding

Cybersecurity company Island reaches valuation of $6.4 billion (Reuters, WHTC). OpenEvidence, dubbed "ChatGPT for doctors," has raised a new round of funding at a valuation of $15 billion (Business Insider). Spanish AI startup Reply Next has raised €400,000 pre-seed (Dealroom). Cyera raised $400m from Goldman Sachs (FinTech Global, supra).

Trends in financial regulation and supervision

A review of the current legal system regarding AI financial services has been published (Center for Democracy and Technology). New York state may be able to share Frontier AI safety reports across the United States without new legislation, Lawfare reported. The EU Securities and Markets Authority (ESMA) has put AI and tokenization on the regulator’s agenda for 2027, reports unchainedcrypto. In Japan, the Australian government revealed that AI had illegally infiltrated the public health insurance system, and that an AI developed by OpenAI was involved, and it was reported that notification to the government was delayed by three months (TBS NEWS DIG).

Integrated analysis: What is happening now

Future points of interest

- To what extent will the assessment conflicts between insurance companies and medical institutions' AIs affect medical costs and insurance premiums? - How far will AI agent-based payments (Visa, Revolut, Kalshi, etc.) expand across major payment networks? - How will the New York State/EU AI safety reporting framework be implemented in AI supervision of financial institutions?