Issue 13
Major U.S. banks block AI agents that log in with customers' credentials
Several major U.S. banks are blocking AI agents that access accounts using a customer's login credentials, saying their terms require human access. Bank of America has asked customers not to give their banking passwords to AI assistants.
Who is affected and how
- retail customer
- Banking passwords and login details should not be given to AI assistants, because the bank prohibits it and the activity is hard to trace.
- banking professional
- Banks already struggle to tell whether the account owner or an AI agent is carrying out an action, so identifying agent traffic is an urgent task.
- fintech company
- Agent products built on sharing customers' credentials rest on a risky business model; official API access is needed.
Bank of England and FCA publish joint report on AI in UK financial services
The Bank of England and the FCA have published a joint report on the role of AI in UK financial services. According to the related survey, oversight of third-party AI systems is the main governance…
banking professionalWhen adopting agentic AI, documenting vendor dependence and governance may become an expectation.layoffs · United KingdomBarclays to cut about 5,000 jobs as it leans on AI
UK bank Barclays said it will cut roughly 5,000 jobs as artificial intelligence and technology take a larger role in its operations. According to the Hungarian report, that is about 5 percent of its…
banking professionalAutomatable back- and middle-office roles are the targets of the cuts, so it is worth retraining toward supervisory and exception-handling skills.chief financial officerFor cost-cutting plans, headcount reductions built on AI efficiency gains now serve as a comparable benchmark.investment · United StatesGoldman Sachs and JPMorgan take leading roles in Anthropic's IPO
Anthropic filed its IPO documents on Monday, which according to reports show rapid revenue growth, and Goldman Sachs is a key player in the offering. JPMorgan is reportedly set to lead the U.S…
banking professionalIn investment and private banking, retail allocation and the handling of AI exposure bring new client questions.chief financial officerThe offering provides a reference point for valuing AI companies and for risk disclosure.fraud · European UnionFraudsters use AI voice imitation to steal huge sum from Italian bank
According to reports, fraudsters used an AI-imitated voice to deceive an Italian bank, and Milan prosecutors are trying to identify the suspected perpetrators. Hungarian outlets report losses of…
banking professionalVoice-based orders and executive instructions require independent, multichannel confirmation.chief financial officerFor high-value transfers, a telephone instruction alone should not be sufficient.Hungary
More key stories of the week
Canada's big banks, including RBC, BMO and TD, block third-party AI agents
Several major Canadian banks, including RBC, BMO and TD, are restricting third-party AI agents, such as Meta Muse, from accessing customer accounts. RBC has barred personal AI agents from its online…
retail customerDo not give out banking credentials to use a third-party AI agent for banking tasks, as the bank may block access.banking professionalDetecting and blocking agent access, and communicating with customers, are new tasks for digital channels.fintech companyAgent-based services must expect banks to block access and should seek official integration.case · globalAnthropic says no single insurer may have capacity to cover a catastrophic AI incident
Anthropic's IPO filing warns that its insurance may not provide sufficient cover for a large-scale AI-related incident, and that no single insurer may have the capacity to do so. Trade press reads…
insurerIn D&O, professional liability and cyber portfolios covering AI providers, accumulation exposure and exclusions need to be re-examined.chief financial officerCompanies using AI need to check whether their existing cover applies to AI-caused losses at all.regulation · United KingdomFCA builds agentic finance rules without introducing AI-specific regulation
The FCA has issued guidance on firms' use of AI and called on AI providers to clearly label financial content. The regulator still will not introduce AI-specific rules, which also raises testing…
banking professionalTesting and quality-assurance procedures for agentic systems need to be tightened.fintech companyAI that produces financial content needs clear warnings and liability frameworks.deployment · European UnionDeutsche Bank uses AI to cut some corporate loan decisions from weeks to minutes
Deutsche Bank has reportedly cut some corporate lending decisions from weeks to minutes by using AI. Hungarian reports say the decision also requires the AI's approval.
banking professionalIn credit assessment, the role of AI and the rules on human approval are part of the competition on turnaround time.chief financial officerFor corporate loan applications, faster decisions can offer an advantage in financing planning.deployment · United StatesShopify adopts Visa, Mastercard and American Express agentic payment technology for AI shopping agents
Shopify is adopting agentic payment technology from Visa, Mastercard and American Express in Shop Pay, while Mastercard is pushing for tokenized card versions for agents. Amex has published an…
fintech companyIn agentic checkout, alignment with network tokens and intent authentication may become a condition of competing.banking professionalIssuers need to prepare to authenticate and fraud-screen the tokenized cards that agents use.case · globalAI liability becomes more pressing as insurers and companies rethink coverage
Chubb's CEO says insurers must address emerging AI liability risks, and trade materials advise companies to review their policies. Terms for liability insurance covering AI agents are still taking…
insurerPricing and exclusions for AI liability products can draw on cyber insurance experience and on questionnaires about third-party tools.chief financial officerA company using AI needs to assess which AI-caused losses its current cyber, liability and professional cover would pay for.survey · globalMoody's and BIS flag private credit and financial stability risks as AI-related debt grows
Moody's says borrowing for AI infrastructure is expanding the private credit market, but near-term maturities could make refinancing harder for weaker borrowers. The BIS flags financial stability…
banking professionalWhen assessing AI clients, lenders should examine business and financial viability; an AI label is not enough.chief financial officerRefinancing risk in AI-exposed loans needs to be monitored.deployment · United StatesLemonade and Insurify launch agentic insurance platforms for AI agents
Lemonade has launched an agentic AI insurance platform for personal AI agents, and Insurify has introduced an AI Agent Gateway that adds consumer protections to AI-based insurance shopping. Analysts…
insurerInsurers need to prepare their channels for quote requests made by AI agents while keeping human experts reachable.fintech companyAgent-gateway models open a new distribution channel in insurance intermediation.tool · globalMcKinsey: agentic AI could cut banks' transaction monitoring workload by up to half
McKinsey says agentic AI could cut banks' transaction monitoring workload by up to 50%. Another analysis summarises one year of experience with AI agents in financial-crime prevention and five…
banking professionalThe work of transaction monitoring analysts is shifting towards reviewing agents' outputs.case · globalBanks hand more decisions to AI as governance lags behind
Analyses say banks are entrusting more decisions to AI while governance frameworks lag, and more banks are worried about AI-related cyber risk. Executives at JPMorgan and Goldman Sachs stressed…
banking professionalThe balance between risk and innovation must also be applied when launching products.survey · European UnionOne in three Europeans would rather discuss sensitive financial matters with AI
A European survey finds that one in three Europeans would rather discuss sensitive financial matters with AI. Some 46% of European consumers say AI has improved communication with their bank, while…
banking professionalOn sensitive topics such as payment difficulties, an AI channel can be customer-friendly if a handover to a human is available.retail customerAn AI channel can be used in case of payment difficulties, but the option of speaking to a human must remain.survey · globalAI models got 57% of financial advice wrong in Saturn test
The Saturn test found AI models erred in 57 percent of cases on financial advice, and according to FT Adviser, losses from bad advice can be as severe as those from fraud. Other industry material…
retail customerFor financial decisions, an AI chatbot's answer should be verified against an independent source.