Tag: fraud

  • Banks Warn AI Shopping Bots Risk Wrong Purchases and Payment Data Exposure

    Banks Warn AI Shopping Bots Risk Wrong Purchases and Payment Data Exposure

    Major banks are sounding the alarm over the rise of AI shopping assistants that can search for products and even complete purchases on behalf of consumers. The concerns range from incorrect purchases and scams to exposure of payment data and weaker consumer protections, prompting calls for clearer standards around AI-driven commerce.

    AI Agents Could Make Purchases on Your Behalf

    The technology, often described as “agentic commerce,” allows an AI system to act on a shopper’s instructions rather than simply providing recommendations. Depending on the design, an agent could search for a product, select an option, and potentially complete the transaction. A group including NatWest, Bank of America, ING, Capital One, ASB Bank, and Commonwealth Bank of Australia has raised concerns about scams, fraud, and misuse of financial data.

    The warning comes as technology companies including OpenAI, Google, Anthropic, and Meta develop AI agents capable of assisting with or carrying out shopping tasks. The banks said adoption is outpacing existing industry standards and consumer protections.

    UK retailer John Lewis reported that searches originating from AI agents rose to 2.5% this year, up from 0.3% a year earlier, signaling rapid growth in AI-driven shopping.

    Card Details and Payment Risks

    Financial information is another major concern. The banking group highlighted situations in which AI agents could request customers’ card details and enter them directly into websites. There are also concerns that an AI agent could steer a shopper toward payment methods that provide weaker fraud or purchase protections. If a transaction goes wrong, consumers may not know whether the retailer, payment provider, or AI service is responsible.

    Banks Seek Clearer AI Commerce Rules

    The participating banks plan to discuss a set of proposals with policymakers. These include requiring businesses to disclose when an AI agent is involved in a transaction and providing greater transparency about how the technology makes purchasing decisions. The group also called for stronger safeguards around customer data, along with freedom for consumers and merchants to choose which AI-powered commerce services they use. Interoperability between different AI shopping systems was another principle included in the framework.

    As AI moves from recommending products to acting on shoppers’ behalf, the debate increasingly shifts toward who controls these transactions and who takes responsibility when something goes wrong.

  • Amazon Flags Growing AI Scam Danger for India’s New Online Users

    Amazon Flags Growing AI Scam Danger for India’s New Online Users

    India’s rapid digital expansion is exposing millions of first-time online shoppers to increasingly sophisticated artificial intelligence scams. Amazon warns that low consumer awareness now presents a major challenge as e-commerce reaches smaller cities.

    The company views education as equally important as technical protection. Fraudsters can use generative AI to produce convincing identities, forged documents, synthetic reviews, and targeted phishing messages.

    AI Fraud Threat Grows Beyond Major Cities

    Rohan Oommen, Amazon’s vice president for customer and partner trust, says India differs sharply from mature e-commerce markets. The country still adds large numbers of first-time users from tier-2 and tier-3 cities.

    “India has a large number of first-time users entering the digital economy. Awareness is much lower, which makes education a significant part of the solution,” Oommen said.

    In established markets such as the United States, online shopping adoption has largely stabilised. India’s expanding user base creates commercial opportunities but also widens the pool of potential fraud targets.

    New shoppers may have limited experience identifying suspicious links, fake customer service calls, impersonation attempts, or manipulated reviews. These risks increase as fraud tools become cheaper and easier to operate.

    Behavioral Data Strengthens Detection

    Meanwhile, Amazon is using artificial intelligence to identify abuse across its marketplace. The company deploys automated systems, including agentic AI, to conduct continuous audits and detect unusual activity.

    Oommen said behavioural patterns provide stronger fraud signals than submitted documents or images. Bad actors can fabricate identification records, seller profiles, and product information using generative tools.

    “Behaviour is much harder to fake than documents or images,” he said. Amazon’s systems analyse years of seller and customer activity to distinguish legitimate behaviour from coordinated abuse. The models can flag suspicious account changes, review manipulation, unusual transactions, and other high-risk patterns.

    The same technology strengthening criminal tactics can also improve platform monitoring. However, automated detection still requires human oversight, enforcement processes, and cooperation with external authorities.

    Phone Scams Remain Harder to Trace

    Additionally, Amazon’s latest Trustworthy Shopping Experience Report highlights the scale of its fraud enforcement. The company removed nearly 300 million fake reviews globally during the reported period.

    Amazon also removed about 70,000 phishing websites and acted against roughly 14,000 phone numbers linked to scams. Oommen said phone-based fraud presents a tougher enforcement challenge than fraudulent websites.

    Investigators can often identify website infrastructure, hosting records, and digital links more easily. Phone scams create attribution problems, especially when criminals use spoofed numbers or rapidly changing communication channels.

    Amazon says it does not wait for new regulations before acting against fraud. Its response includes automated detection, account enforcement, consumer education, and cooperation with law enforcement agencies.