Issue 11
JPMorgan blocks outside AI agents from customer accounts; Schwab also blocks
JPMorgan Chase chief executive Jamie Dimon said on October 12–13 that the bank will not let outside AI agents, including Muse and Instinct, move money or pay bills from customer accounts. Schwab is also blocking unauthorized agents, and reports say liability for an agent's actions could fall on the customer.
Who is affected and how
- retail customer
- Do not give an AI agent access to your account: if you do, every step the agent takes may be your responsibility.
- banking professional
- Identifying and blocking agent traffic, and handling customer consent, should be written into policy.
- fintech company
- A provider offering an outside agent will have no direct entry to large-bank accounts and must negotiate an authorized access route.
Bank regulators in the US and Europe tighten scrutiny of AI risks
US and European bank regulators are stepping up checks on AI risks at the same time, according to reports on October 13. Banks are being asked about their use of Chinese technology and about…
banking professionalCompile an inventory of AI use, including components of Chinese origin, and have a documented shutdown mechanism for every autonomous agent.chief financial officerExpect stricter supervisory scrutiny of substitution and outsourcing risks in AI systems.regulation · globalFSB's final report sets out 12 sound practices for responsible AI adoption
The Financial Stability Board published its final report on October 15, setting out 12 sound practices for the responsible use of AI in finance. According to Reuters, the body calls for stricter…
banking professionalCompare the human-oversight points in your agentic AI processes against the 12 practices.insurerThe practices are also worth using as a reference when deploying AI agents in insurance processes.investment · globalAI financing tightens as issuance falls and default protection gets costlier
Reports say AI-related debt issuance fell sharply in September, data-center bond prices dropped, and credit insurance on the large cloud providers became more expensive. Central bankers are weighing…
chief financial officerReview AI-linked loan and bond exposure and expect more expensive refinancing.banking professionalTreat AI-infrastructure customers separately within credit risk limits.insurerThe rising cost of credit insurance is a risk-pricing signal for the cloud giants; build it into portfolio analysis.supervision · globalIMF stability report flags AI market correction, rising yields and cyber risk
The IMF's Global Financial Stability Report, published on October 13, says the resilience of financial markets is no guarantee of safety. An AI-driven market correction, rising yields and cyber…
chief financial officerRun a stress scenario for an AI correction and higher yields in funding and investment plans.banking professionalCheck AI-related credit and market exposure.insurerReview the concentration of AI-linked assets in investment portfolios.Hungary
More key stories of the week
Big banks acknowledge AI financing risks but keep lending to the boom
According to one report, large US banks are financing AI investments despite an exposure of about $800 billion, while risks grow on both sides. JPMorgan and Goldman Sachs, however, are cautious about…
banking professionalIn syndicating AI loans, the large banks' caution can serve as a benchmark for pricing and risk-taking decisions.chief financial officerFor loans raised for AI infrastructure, expect high yields and narrowing investor appetite.survey · globalChainalysis: AI agents made 138 million crypto payments in a year, most of them tiny
According to a Chainalysis report, payments made over Coinbase's x402 protocol grew from about 1,500 in mid-2025 to more than 138 million by the third quarter of 2026. Most payments are small, and…
fintech companyMachine-to-machine micropayments point to a real market, running on stablecoin rails.chief financial officerWhen buying agent services, plan for stablecoin payouts and a control system.fraud · globalBanks seek customer trust against AI voice cloning as SBI warns and call protection spreads
The State Bank of India (SBI) warned on 17 October about AI voice-cloning scams and recommended five ways to protect against them. A trade article says banks must call customers in a way that…
retail customerDo not transfer money at the request of an unknown voice; call the bank back on its official number.banking professionalCustomer calls must be made verifiable and recognisable.supervision · globalFSB chair names AI and stablecoins as priorities for the G20
The FSB chair set out the board's work on AI and stablecoins for the G20 on October 12, led by Andrew Bailey according to reports. Bailey also warned of cyber threats to the banking system, referring…
fintech companyProviders of stablecoin-based and AI-agent services should prepare for tighter compliance expectations.oversight · AsiaPing An Digital Bank reaches 96% deepfake detection accuracy in HKMA GenAI sandbox
Ping An Digital Bank said on 16 October that it detected deepfakes with 96% accuracy in the Hong Kong Monetary Authority's (HKMA) GenAI sandbox. The Hong Kong banking association welcomed the HKMA's…
banking professionalThe sandbox result is a benchmark for the accuracy of deepfake-detection systems.survey · globalReport: AI could free up 40% of banks' middle and back-office capacity by 2030
A report dated 14 October says AI could free up 40% of banks' middle and back-office capacity by 2030. Banks are competing with big tech for AI-skilled staff, and Bloomberg reports that junior…
banking professionalEntry-level tasks are shifting toward supervising AI agents, so it is worth building skills for that.chief financial officerWhen planning for freed-up capacity, factor in the pay competition for AI talent.survey · globalCFO surveys: AI use is growing, but returns and trust remain uncertain
A OneStream index finds that finance is scaling AI faster than it can trust the results, while Tipalti says AI adoption has tripled but teams are unprepared. Another report says half of CFOs consider…
chief financial officerBefore starting AI projects, set return metrics and checkpoints.accountantVerifying AI-generated financial results may become the next skill requirement.case · globalAI agents challenge payment intent verification and fraud liability
Between 12 and 15 October, several analyses warned that software buying on a person's behalf blurs the line between fraud and consent. Mastercard is preparing for "rogue" agents going on spending…
banking professionalDevelop intent and consent checks for transactions made by agents acting on a customer's behalf.chief financial officerWhen granting agents payment authority, account for the costs of identity verification, oversight and liability.fintech companyBuild verifiable intent capture and spending limits into your payment product.supervision · United KingdomFCA to automate supervision with AI, with no new rules for chatbots
The UK's Financial Conduct Authority (FCA) said on October 14–15 that it will automate supervision with AI and, for now, will not create new rules for chatbots. The AI Live Testing service launched…
banking professionalCompliance is judged under existing rules rather than new chatbot rules, so it is worth documenting internal controls.fintech companyAI Live Testing remains a way to trial AI products.regulation · United StatesAICPA issues practice aid on using AI in financial statement audits
The AICPA published a practice aid on October 13 for auditors on using AI in financial statement audits. The document gives guidance on AI-assisted auditing.
accountantBefore using an AI tool in audit work, compare your process with the AICPA practice aid.chief financial officerExpect your auditor to use AI, and ask about controls.case · globalInsurers weigh AI exclusions as Marsh chief warns of uncertainty
On 15–16 October, specialty insurers weighed AI liability risks and resisted exclusions, while Marsh's chief executive said broad AI exclusions in policies could create uncertainty. Insurance has…
insurerThe wording of AI exclusions determines which losses are covered, so clarify the terms.chief financial officerCheck whether your company's insurance contains an AI exclusion.deployment · United StatesAmerican Express publishes its standard for AI agent payments
On 14–15 October, American Express opened up its rules for AI agent payments. According to reports, the standard declines unauthorised AI purchases before checkout is completed.
fintech companyIf you build an agent payment product and expect card acceptance, align with the AmEx standard.banking professionalPre-payment agent verification offers a model for reshaping fraud screening.