AI & Finance News — 2026-10-22 (Thu)

Overview

The past 12 hours were dominated by a sharpening AI bubble debate, with Citadel's Ken Griffin saying investors are "almost certainly" in the midst of one (The Hill) and BlackRock's CEO warning of trouble brewing in a $1 trillion AI boom (TradingView). Operational risk from AI also surfaced on both sides: JPMorgan will not let outside AI agents into customer accounts (ABC Money), and a suspected AI-driven system hijack produced unauthorized trades at Japan's Iwai Cosmo Securities (San-in Chuo Shimbun, Hokkoku Shimbun). Wall Street still closed higher as a strong SpaceX profit lifted AI stocks (Yahoo Finance).

Banking

Commentary argued that when banking AI meets cloud failure, "fail safe" needs to become a design principle (Digital Journal), while the effectiveness of banks' AI chatbots for customer service remains in question (WKYC). AI security specialists on $200k salaries in banks could earn 4x more by leaving, highlighting retention pressure (eFinancialCareers). In Japan, mega banks and online banks are developing IT in India, with Mitsubishi UFJ using AI tools in Bengaluru (Nikkei). Police arrested two suspects in an AI-powered bank fraud that cost victims €380,000 (NL Times), and Iwai Cosmo Securities reported trades settled in 5 accounts that customers did not intend, with a system hijack and possible AI use suspected (San-in Chuo Shimbun, Hokkoku Shimbun, JoongAng Ilbo).

Insurance

A claims tribunal was urged to address a self-represented claimant's use of AI tools in a health insurer dispute (Global Competition Review). Patra appointed Yousef Hashimi to lead AI products and client transformation (citybiz). AI skills demand and retirements will keep insurance talent scarce through 2027 (Insurance Business). Micro enterprises underestimate their appeal to cyberattackers as AI phishing grows (Life Insurance International). 【Asset management and market trading】 BlackRock added a new robotics ETF, with exposure to Nvidia, Tesla and SpaceX, to its growing lineup of AI-themed funds (Mezha.net, TradingView), expanding its AI push even as investor enthusiasm cools (The Economic Times). Asset management executives said the investment boom could last another three to four years (Futu). The Nasdaq led Wall Street to a higher close as SpaceX and Super Micro boosted AI sentiment (Yahoo Finance), with a strong SpaceX profit report helping lift AI stocks and Wall Street (GMToday.com). CNBC's "Fast Money" traders discussed CoreWeave's stalled bounce and risks in the AI trade (CNBC), and the WSJ's Morning Risk Report flagged AI running into its next big obstacle (WSJ).

Fintech and payments

Agentic commerce took center stage on day two of Money20/20 USA 2026 (FinTech Futures). Stripe launched Stripe Treasury to manage finances using AI agents (TradingView), and Mastercard began offering early access to an AI agent payment service (NADA NEWS). Evervault announced an AI Token Service (Digital Transactions). Six hard truths are stalling AI in financial crime compliance (FinTech Global). 【Statements by financial leaders and officials】 Jamie Dimon, JPMorgan Chase CEO, said JPMorgan will not let outside AI agents into customer accounts, as banking asks who gives them a mandate (ABC Money). Ken Griffin, Citadel CEO, called AI a bubble after Citadel bought the dip (Benzinga), saying investors are "almost certainly" in the midst of an AI bubble (The Hill), in what was described as a stark call on the AI boom (TradingView). Larry Fink, BlackRock CEO, sees trouble brewing in a $1 trillion AI boom (TradingView). Finance leaders sparred over the AI bubble after Griffin's warning (AdvisorHub). 【AI-related stocks and fundraising】 Wall Street ticked higher after SpaceX helped lift AI stocks (WTOP News), while Treasury yields dipped as AI stocks tested investors' nerves (Finimize). Legal AI startup Legora landed $550M at an $8.8B valuation (The Tech Buzz). Peak XV-backed AI startup Blacksmith netted $66 Mn at a unicorn valuation (Inc42). 【Financial regulation and supervision】 The Payments Association issued quick comments on the FCA's AI regulation ambitions (Insurance Edge). UK financial services firms must adapt to an age of regulatory observability as AI spreads (Pinsent Masons). 【Integrated analysis: what is happening】 Financial institutions are moving AI from experiments into products — a robotics ETF (TradingView), AI agent payment services (NADA NEWS), AI-agent-managed treasury (TradingView) — while funding continues at scale, with $550M for Legora (The Tech Buzz) and $66 Mn for Blacksmith (Inc42). At the same time, risk is turning concrete on both sides of the technology: suspected AI-driven attacks hit a Japanese securities firm (Hokkoku Shimbun) and bank fraud cost victims €380,000 (NL Times), while compliance failures and talent shortages are cited as barriers in insurance and financial crime work (Insurance Business, FinTech Global). The bubble dispute (The Hill, AdvisorHub) suggests capital may become more selective even as product launches continue; a robotics ETF debuting as investor enthusiasm cools (The Economic Times) captures that tension. Regulation appears to be trending toward closer observability of firms' AI use, with the FCA's ambitions already drawing industry response (Insurance Edge, Pinsent Masons).

What to watch

Whether other banks follow JPMorgan in barring outside AI agents from customer accounts, and who defines the mandate rules (ABC Money). How the bubble dispute translates into fund flows after Griffin's warning and BlackRock's caution, and whether demand for new AI-themed ETFs holds up (The Hill, TradingView, The Economic Times). Any regulatory follow-up in Japan to the suspected AI-driven hijack at Iwai Cosmo Securities (San-in Chuo Shimbun) and the FCA's next steps on AI regulation (Insurance Edge).