Issue 8
Viewpoint
fraud · European Union
Voice-cloning scam at Intesa Sanpaolo leads to €95 million in transfers
Fraudsters used AI to imitate the voices of executives at Italian bank Intesa Sanpaolo in WhatsApp calls and, according to reports, triggered €95 million in transfers abroad. The case was reported on 25–26 September 2026.
Who is affected and how
- banking professional
- Transfer instructions based on executive calls must be confirmed through an independent call-back.
- chief financial officer
- The approval chain for exceptional, urgent transfer requests should be extended with multi-step verification.
fraud · global
Major banks warn that AI shopping bots raise fraud and data-privacy risks
Major banks warn that AI agents and chatbots that shop online carry risks of fraud, misuse and data privacy. The warning ran across international and Hungarian media on 22–23 September 2026, as AI…
retail customerGive card details or payment authority only to a verified AI agent with limited scope, and monitor its transactions.banking professionalCustomer communication and fraud prevention need a separate scenario for unauthorized payments initiated by an agent.deployment · European UnionDanske Bank and Mastercard complete Denmark's first agentic AI payment; Rogers Bank transaction in Canada
Danske Bank and Mastercard completed Denmark's first agentic AI payment. An agentic commerce transaction involving Mastercard, Flybits and Rogers Bank was also completed in Canada, according to…
banking professionalAgent-initiated transactions are already running live, so issuer-side handling and authentication should be tested.fintech companyIntegration with card networks' agent payment frameworks can be a competitive advantage.chief financial officerCompanies may soon receive payments initiated by AI agents, so approval frameworks need to be set up.supervision · European UnionESMA makes AI and tokenisation an EU supervisory priority for 2027
From 2027 the European Securities and Markets Authority (ESMA) will make digital innovation a supervisory priority, initially focusing on financial firms' use of AI and on tokenisation. The authority…
fintech companyFirms using tokenisation and AI should expect increased supervisory attention.survey · globalReport: AI chatbots mostly give wrong financial advice
A new report finds that AI answers financial advice questions incorrectly in most cases. One article says losses from wrong AI advice could be as severe as those from scams (21 September 2021).
retail customerFinancial decisions should not be made on an AI chatbot's answer without verification.banking professionalBanks' own chatbots should be tested for the accuracy of their financial answers.Hungary
deploymentK&H prepares to accept payments initiated with Mastercard Agent Paybanking professionalAgent-initiated payments are also appearing at Hungarian banks, and the process for identifying and handling them must be developed.fintech companyK&H's move signals that agent payment acceptance capabilities are also being built in the domestic market.retail customerA customer's AI assistant may in future initiate purchases through the bank, so it is worth clarifying the approval conditions.caseAI could bring cost cuts and profit pressure to Hungary's financial sectorbanking professionalBanks should expect AI adoption to affect costs and headcount, while addressing regulatory concerns.insurerHungarian insurers' operations also show efficiency-oriented AI use, alongside regulatory questions.chief financial officerAI investment plans should weigh the combined effect of lower costs and profit pressure.surveyGránit Bank survey: bank customers more open to AI financial helpbanking professionalCustomer openness supports rolling out AI assistants, but trust and error handling are preconditions.retail customerAs AI support spreads, customers should consider carefully which financial decisions they hand over to it.investmentAI accounting: Integral raises €18m, Tabby serves 5,500 small businesses, Hungarian Cashbook conferenceaccountantAutomation of bookkeeping, tax and payroll is changing the content of the service and the client relationship.fintech companyAI accounting platforms are attracting capital and reaching small-business customers.chief financial officerAgentic accounting solutions are already available for small businesses.caseHungarian industry forums: who decides, who is accountable, who stops live bank AIbanking professionalDecision, accountability and shutdown powers must be clarified before going live.insurerIn claims handling, too, the question has arisen of where a human should step in during AI-supported customer service.investmentFintechOS raises $28M after reaching profitability; Creatio to invest $300M in agentic AIfintech companyAI software vendors supplying banks and insurers are gaining capital and expanding.banking professionalThe range of vendors is widening, but buying agent-based solutions raises governance questions.
More key stories of the week
investment · global
BlackRock: AI agents could drive stablecoin demand in machine-to-machine payments
BlackRock says the small, continuous payments AI agents make for data and computing capacity could run on stablecoins and blockchain-based payment systems, because card networks and ACH fit…
fintech companyMachine-to-machine micropayments could open a new product area for stablecoin-based infrastructure.banking professionalReplacing card and ACH rails for machine payments is being raised as a possibility, so strategy should account for it.chief financial officerThe scenario is so far a forecast; actual agent payment volumes are still small.deployment · United StatesAI-native insurance players: Corridor brokerage, MGT's California expansion, Athenium CairnQA
Corridor launched an AI-native benefits brokerage for small businesses, AI-based insurer MGT extended its small commercial coverage to California, and Athenium launched CairnQA for claims and rating…
insurerAI-native players are creating new quality-assurance and competitive pressure in claims handling and underwriting.fintech companyAI-based insurance models are expanding in the small-business segment.deployment · globalNew payment rails for AI agents: Polygon, Coinbase x402, Ripple MPP, Mesta, Kredete
Several providers announced agent payment infrastructure: Polygon is optimising for agent payments, Coinbase offers x402, Ripple has joined MPP, Mesta has launched agentic payments, and Kredete has…
fintech companyStablecoin-based agent payment rails are multiplying quickly, so the choice of standard is a strategic question.banking professionalThe spread of payment rails outside the banking system challenges the traditional role of banks in payments.tool · globalVisa, Mastercard and Ant build shared Know-Your-Agent framework as Sumsub links KYC to agent transactions
Visa, Mastercard and Ant are working on a shared "Know-Your-Agent" framework, while Sumsub and Sumvin are linking KYC identity checks to AI agent transactions. Trade coverage says banks need to…
banking professionalIdentifying the agent and proving human approval add a new authentication layer to the payment flow.fintech companyJoining shared identification frameworks may become a condition of distribution.case · United KingdomAudit and AI: Modus hires ex-Grant Thornton auditors, PwC sets human-oversight guardrails
Modus has hired former Grant Thornton auditors for its AI-native model, and PwC is limiting the use of AI in audit through human oversight. The Big Four are also rethinking graduate training because…
accountantAI-native models and changing Big Four practice are reshaping entry-level tasks and the training path.chief financial officerAuditors' use of AI may show up in the audit process and in fees.case · United KingdomCustomer-service chatbots: UK told to keep human access; wrong-advice and insulting-reply cases
UK banks and energy firms have been told not to rely solely on AI chatbots for customer care. One chatbot wrongly told a mortgage borrower that a payment holiday would not affect their credit score…
banking professionalAccess to a human agent must always be provided alongside the AI chatbot.retail customerOn an important lending question, confirm the chatbot's answer with a written, official source.survey · globalGartner: CFOs must be more disciplined about AI expectations
Gartner research says CFOs need to manage AI expectations more realistically. Professional articles also cover accounting governance for AI agents and practical steps for tax leaders (24–25 September…
chief financial officerReturn expectations for AI investments should be managed with discipline and measurable targets.accountantGovernance rules are needed to build AI agents into the accounting process.investment · United StatesJPMorgan's Dimon: AI infrastructure spending could approach $1 trillion in 2027
JPMorgan CEO Jamie Dimon said spending on AI infrastructure could approach $1 trillion in 2027. The statement was published on 21–22 September 2026.
banking professionalThe financing volume of AI infrastructure raises questions of credit and risk exposure.chief financial officerIn capital and credit markets, the scale of AI investment may affect financing conditions.regulation · globalGlobal banks work together on principles for agentic commerce
Major global banks have aligned their positions and are mapping out principles for agentic commerce. The news appeared on 22 September 2026.
banking professionalThe emerging industry principles may shape internal rules for agentic payments.oversight · AsiaSingapore's MAS reviews deepfake scam defences as cross-bank AI shield is tested
The Monetary Authority of Singapore is reviewing its defences against deepfake scams after a loss of nearly $4 million, while testing an AI protection system shared between banks. The reports…
banking professionalBecause of deepfake-based fraud, authentication and confirmation processes need to be reviewed.layoffs · European UnionDanske Bank expects further job cuts as AI use expands
Danske Bank's workforce is expected to shrink further as its use of AI grows, according to reports. The news was reported on 21–22 September 2026.
banking professionalAI-driven efficiency gains are having a concrete headcount impact at a large Nordic bank.chief financial officerIn cost-reduction plans, the link between headcount and AI investment must be modelled.investment · globalOracle's Project Jupiter loans fall below 90 cents as SoftBank launches $10 billion-plus bond sale for OpenAI stake
Oracle's Project Jupiter loans have fallen below 90 cents on the dollar, SoftBank Group is launching a bond sale of more than $10 billion to fund its OpenAI investment, and Bank of America is lagging…
banking professionalTaking part in AI debt deals, or missing out on them, affects trading and investment banking revenue.chief financial officerSecondary-market pricing of AI loans can signal rising financing costs.In brief
toolFeedzai launches Farol, an agentic AI for fraud investigation that cuts investigation timesbanking professionalAutomating fraud-investigation queues may reduce analysts' manual workload, but responsibility for decisions and auditability need clarifying.fintech companyFraud-analytics agents are becoming a competitive product category.oversightFCA Mills Review AI findings; Claude for financial advisers with Schwab integrationbanking professionalResponsibility for deploying AI assistants built into advisory workflows needs to be clarified.fintech companyAdvisory AI integrations are now appearing at large platforms too.caseInsurers split over AI coverage as CFC launches IP cover and Munich Re's aiSure backs live lending modelinsurerCovering or excluding AI risks requires a product-level decision.chief financial officerCompanies that use AI should check which AI risks their existing policies cover.oversightFrontier AI is testing operational resilience in European bankingbanking professionalOperational dependencies and failure handling of AI systems must be built into resilience plans.fraudAI-driven fraud wave: insurer complaints teams overloaded, Irish bank liability debate, deepfake investment scams in KazakhstaninsurerWithout a process to filter genuine from AI-generated complaints, legitimate cases may face delays.banking professionalThe question of bank liability to victims of deepfake fraud is a legal and customer-handling risk.oversightFed's top bank supervisor focuses on banks protecting themselves from AI risksbanking professionalUS supervisors are paying increased attention to how banks manage AI risks internally.regulationAI supervision frameworks: US state regulators, CFTC forum, Bank of England deputy governor's warningbanking professionalHuman oversight of autonomous AI systems needs to be built into documentable processes.fraudAI-forged receipts and supplier payment fraud: controls for finance teamschief financial officerChanges to supplier bank details and payment approval should be tied to a human, two-step control.accountantSubmitted receipts and approvals must be handled with authenticity checks, given the possibility of AI forgery.caseMeta AI agent triggers selloff in bank, insurer and travel stocks; Insurify blocks Meta's Muse agentinsurerConsumer AI agents may displace or reshape comparison and intermediary channels, so handling agent traffic is a strategic question.fintech companyComparison sites must decide whether to serve third-party AI agents.regulationAICPA and former IRS commissioner Danny Werfel launch initiative on AI tax risksaccountantIndustry guidance on applying AI in tax is being prepared and is worth following.chief financial officerResponsible-use practices for AI in tax should be documented.fraudAI attack reportedly yields data of more than 600,000 bank cardsbanking professionalWhere card data is affected, issuers must check for signs of misuse and replace cards where necessary.retail customerAfter news of a card data leak, it is worth monitoring transactions and setting up alerts.oversightS&P: AI use will soon influence banks' credit ratingsbanking professionalAI strategy and risk management will become factors in credit rating assessments.chief financial officerA bank's AI adoption and governance data will also need to be communicated to rating agencies in a structured way.regulationOCC conditionally approves bank charters for AI-agent and stablecoin infrastructurefintech companyThe US charter route may open up for agent and stablecoin infrastructure as well.banking professionalNew licensed competitors may enter agent payments and stablecoin services.fraudMicrosoft shuts down EvilTokens, an AI-based phishing servicechief financial officerInvoice fraud originating from compromised mailboxes remains a threat to prepare for, for example by independently confirming changes to bank account details.accountantUnexpected requests from clients to change payment or invoice details must be verified through a separate channel.banking professionalPayment-fraud patterns may also rely on AI-based analysis of compromised email accounts.surveyHalf of prop trading firms let staff use AI freely while banks set limits; AI-run funds emergebanking professionalTighter AI limits at banks may mean a competitive disadvantage or lower risk compared with trading firms.fintech companyAI-run investment funds are appearing as products.deploymentBill Harris launches free AI financial planner Evergreen.airetail customerFree AI-based financial planning is available, but caution is warranted about the accuracy of AI advice.fintech companyA well-known founder entering consumer AI financial advice means a new competitor.deploymentUK-backed AI system strengthens Pakistan's FBR tax monitoring in the cement sectorchief financial officerAI-based data analysis by tax authorities raises compliance and data-quality expectations.accountantWith AI-based tax-authority checks, consistency of filing data matters even more.surveyTemenos and Celent: banks face a fight to keep three in four customers as expectations rise in the AI erabanking professionalCustomer retention depends on improving AI-based customer experience and on how a bank's reputation appears through AI.fintech companyRising customer expectations are pushing the market toward digital and AI-based services.toolContent-provenance verification could help against deepfakes and fake newsbanking professionalVerifying the origin of content could add protection to customer and executive identification.regulationBrazil tightens Pix rules over deepfake fraud; South Korea overhauls credit information consentfintech companyThe overhaul of the consent system gives room to develop AI assistants in South Korea.banking professionalDeepfake-related rules for instant payment systems could serve as a model elsewhere.caseWar risk and insurance gaps threaten US data centres being built in the Middle EastinsurerCovering war risk for data centres, and the insurance gap, is a question of risk and pricing.banking professionalThe resilience of banking services that depend on AI data centres must be assessed.