Skip to content

All updates

Today 1 items
Oct 14Wed
  1. PYMNTSSource summary46

    Emirates NBD, Mastercard and Revolut bring AI travel planning into payment decisions

    Emirates NBD, Mastercard and Revolut are embedding AI travel planning services into banking or travel platforms to influence decisions before consumers make payments. Emirates NBD, Visa and Tripadvisor have launched a free assistant that does not require a bank account, can generate itineraries based on preferences, budgets and styles, and displays offers for Visa cardholders.

  2. PYMNTSSource summary45

    61 million U.S. consumers have started using AI assistants for shopping searches

    A PYMNTS Intelligence report shows that nearly 61 million U.S. consumers now begin their online retail research with AI assistants, as AI accelerates changes in product discovery. AI-assisted shoppers’ average quarterly spending rose 17% in August to $1,039, indicating that they have put conversational tools to use for actual purchases. 59% of AI shoppers now research on AI platforms products they previously searched for on Amazon, while 45% consequently discovered online stores they had never visited.

  3. PYMNTSSource summary26

    Sereact、Zalando and CEVA Logistics deploy AI robots for returns sorting in Germany and Poland

    Sereact、Zalando and CEVA Logistics have deployed AI robots in returns operations in Germany and Poland, enabling the identification, picking and sorting of apparel without item-by-item training. The system runs on Sereact’s Cortex platform and is delivered as Robotics-as-a-Service; UPS-owned Happy Returns has increased sorting efficiency by 50% using technologies including Geekplus robots.

  4. Banking DiveSource summary59

    JPMorgan CEO Dimon warns customers that using banking agents means assuming liability

    JPMorgan CEO Jamie Dimon warned that customers who allow AI agents to access their bank accounts will need to bear the liability risks arising from that decision. JPMorgan will not currently allow arbitrary external agents to connect to its internal systems and is building related safeguards; its AI-powered Smart Cash remains in the pilot stage and can transfer customer funds between accounts at the bank to improve returns and manage cash flow.

  5. Stripe BlogSource summary26

    The “Alien” Economics of AI Monetization

    Among the companies on the Forbes AI 50 list, two-thirds have adopted some form of usage-based pricing. AI companies are differentiating themselves from traditional SaaS through faster pricing iteration and more diversified market strategies. Fin AI Agent adopts a tiered model combining subscriptions, usage-based pricing, and credits, while Replit combines subscriptions, usage-based pricing, and rate-limited Free Mode to respond to changes in costs and usage.

Oct 13Tue
  1. Digital Transactions NewsSource summary71

    AmEx launches an agentic commerce framework, and Forter introduces Agentic Commerce Engine

    AmEx launches Intent Intelligence Framework, which captures and authenticates customer intent before an agent completes a purchase, providing it to issuers for payment authorization, fraud prevention, and dispute resolution. Forter introduces Agentic Commerce Engine to help merchants verify agent shoppers and manage agent transaction processes, fraud risks, and activity monitoring.

    推荐理由:AmEx brings customer purchase intent into the payment process earlier and connects it with authorization and risk controls, while Forter adds agent verification and transaction management capabilities from the merchant side.

  2. PYMNTSSource summary48

    Entrust: Millions of AI agents will test enterprises’ ability to control them

    Entrust Chief Operating Officer Anudeep Parhar said enterprises may deploy millions of AI agents in the future, making authorization and accountability critical to scaling agentic commerce. Entrust believes authentication can only determine whether an agent can access a system, while runtime authorization must also determine whether it is permitted to perform a specific action. Enterprises need to make high-impact operations traceable to the person who authorized them and retain evidence to address audits and regulatory requirements.

  3. CryptopolitanSource summary21

    Bitwise Chief Investment Officer Matt Hougan: NEAR Could Rise 40x by 2030 in an Optimistic Base-Case Scenario

    Bitwise Chief Investment Officer Matt Hougan believes NEAR could rise to 40 times its current value by 2030 in an optimistic base-case scenario, with a target price of $155.85. NEAR Intents has processed nearly $30 billion in cross-chain transactions to date, while payment and trading needs from AI agents are seen as potential sources of growth. Bitwise has launched a spot NEAR ETF on NYSE Arca, trading under the ticker NRR, with an annual fee of 0.75% and an estimated annual staking yield of 5.3%.

  4. CryptopolitanSource summary13

    Meta launches two AI models, Muse Wonder and Muse Wonder Full, to challenge ChatGPT and Claude and drive subscription revenue

    Meta launches two AI models, Muse Wonder and Muse Wonder Full, competing with ChatGPT and Claude while seeking to generate revenue through subscriptions. Muse Wonder is available through the free Muse personal agent app, with subscriptions priced at $20 or $100/month; developers can also access it through Meta’s Model API, at a higher cost than the standard subscriptions.

  5. PYMNTSSource summary63

    Meta and Sierra launch Poppy, allowing customers to set AI agent permissions

    Meta and Sierra are building the open standard Personal Agent Protocol (Poppy), allowing AI agents to identify themselves to merchants while letting customers decide whether they can view accounts, modify accounts, or both. Poppy announced 35 design partners; in a demonstration, Meta’s Muse connected to Rocket’s AI agent and received a mortgage pre-approval letter within moments after the user shared credit and income information. Poppy does not yet handle payments, and Sierra has added payment extensions to its roadmap.

Oct 12Mon
  1. PYMNTSSource summary27

    AI agents give small businesses enterprise-level operational capabilities

    OECD says small businesses are restructuring operations around AI agents rather than employee headcount, with customer service, compliance monitoring, and supply chain management now the three leading production use cases. The global AI agents market is expected to grow from $5.4 billion in 2024 to $236 billion by 2034, while agentic AI could generate $3 trillion in enterprise productivity gains over the next decade. HBR recommends setting clear responsibilities, spending limits, and human escalation paths for agents.

  2. Dogpay Blog (中文)Source summary6

    What authorizations, spending limits, and transaction records do AI agents need before initiating payments?

    The article does not specify the authorizations, spending limits, or transaction records required before an AI agent initiates a payment. DogPay App is provided by STRATUSPAY FINTECH LIMITED, registered in Canada, which is regulated by FINTRAC and registered as an MSB, with license number C100000785; its payment products are provided through partnerships with licensed institutions, and the cards are issued by local licensed partners.

Oct 10Sat
  1. PYMNTSSource summary61

    How Banks Evaluate AI Agent Performance

    PYMNTS reviews how BNY, DBS and FINRA manage AI agents in banking, covering performance tracking, human review, knowledge capture and operational audit trails. BNY says its payments agent reduced verification time from up to 6 minutes to under 30 seconds, while outstanding payment investigations fell by nearly 80%, though neither figure has been audited. DBS has made a credit agent available to approximately 1,500 employees globally, while FINRA has urged broker-dealers to focus on agents’ autonomous decision-making, permission boundaries, human approval and audit trails.

Oct 9Fri
  1. Payments DiveSource summary58

    How agentic payments create risks

    The liability boundaries for agentic payments remain unclear. If an AI tool initiates a transaction beyond the scope of a consumer’s authorization, Regulation E and Regulation Z still cannot clearly determine who bears the loss. The article recommends distinguishing between the actions of consumers and AI tools, recording parameters such as the amount, merchant, and time of transaction authorization, and avoiding shifting the allocation of losses to consumers.

  2. PYMNTSSource summary25

    Cost and compliance barriers may slow the adoption of AI insurance comparison agents

    William Blair believes AI insurance comparison agents will change how consumers buy insurance, but cost, regulatory and trust barriers mean they are unlikely to reshape the insurance distribution landscape in the near term. For insurers to obtain a formal, bindable quote, external data costs can reach $20; continuous AI comparison shopping will also drive up customer acquisition costs and create compliance risks involving disclosures and commissions. The report expects the online insurance market to grow through 2027, but at a rate below the approximately 30% level in the first half of 2026, and licensed human agents may still need to participate in the process.

  3. PYMNTSSource summary46

    Splitit CEO: AI agents need complete financial information to recommend Pay Later

    Splitit CEO Nandan Sheth said AI agents need to consider cash flow, existing debt and the total cost of borrowing to determine whether consumers should choose Pay Later. Joint research by Splitit and PYMNTS Intelligence showed that 61% of consumers would accept credit or Pay Later recommendations from AI, but only 2% would let an agent decide on its own.

  4. PYMNTSSource summary15

    Natura hopes users will control AI agents through the Interface smart ring

    AI and hardware startup Natura launched the $99 Interface smart ring, allowing users to interact with AI agents and complete tasks while tracking their health. The ring can be used to record workouts, check schedules, reorder shampoo, turn off the TV, and set alarms. A joint study by Visa Acceptance Solutions found that 56% of consumers allow AI agents to search for and compare products, but only 35% are willing to let them access saved payment credentials.