Google's June 2026 Feature Drop introduced Fake Call Detection within the Phone app. By utilizing Rich Communication Services (RCS) to verify the originating device against the user's contact list, the OS detects caller ID spoofing and flags the discrepancy visually before the call is answered.29
Furthermore, Android 16 introduces systemic blocks for high-risk actions. If a user is on an active phone call with a non-contact number, the OS automatically disables the ability to grant Accessibility permissions to newly downloaded apps, disable Google Play Protect, or sideload APKs.28 This severs the primary technical mechanism of tech-support and "safe account" scams, where fraudsters instruct victims on the phone to download remote-access trojans.
[6] The Economics of Proactive Trust [source]
Fraud prevention is a primary driver of customer retention. Firms that implement modern, proactive fraud prevention programs achieve a 12.3% annual increase in customer retention rates, compared to 7.0% for firms using legacy methods—a 76% performance gap.30 These proactive institutions also record a 94% greater annual improvement in customer satisfaction (CSAT) scores.30
Conversely, reactive recovery destroys loyalty. While 92% of bank customers are likely to reuse their bank if a fraud issue is successfully resolved, the initial breach leaves permanent scars.31 In ecommerce and retail environments, fraud accounts for 31% of all attacks against loyalty programs.32 Because these accounts hold dormant value, they are 4-5 times more likely to be attacked than standard profiles.33 Consequently, 69% of loyalty executives report that fraud negatively impacts brand perception, and 25% of loyalty members will cancel their membership entirely if their account is compromised.33
Proactive engagement flips the dynamic. Customers alerted to potential fraud before funds leave the account view the friction not as an inconvenience, but as a high-value service. Automating these defenses also scales agent efficiency. Contact centers utilizing predictive fraud analytics post a 12.5% annual increase in agent productivity, compared to 2.4% for standard centers.30 Agents spend less time executing rigid identity verification interrogations and more time resolving complex customer needs.
[7] The Agentic Frontier: Know Your Agent (KYA) [source]
By 2026, the primary user of a banking interface is increasingly an autonomous software agent. Forrester estimates that machine-initiated traffic to banking systems will surge by 40% as consumers delegate research, budgeting, and purchasing tasks to personal AI assistants.34
This shift breaks traditional authentication paradigms. Existing fraud models track human keystrokes, tab-switching, and geographic location data. AI agents process transactions via API, execute tasks instantly without typos, and operate headlessly.35 A payment system demanding a human CAPTCHA or a biometric face scan fails when an AI agent attempts to execute a pre-authorized $45 grocery order.
The industry is rapidly developing "Know Your Agent" (KYA) frameworks to solve this authorization gap. KYA explicitly binds a machine identity to a verified human entity, ensuring agents cannot hold or move funds anonymously.36
Leading UX patterns for KYA center on access management rather than point-of-sale friction:
- The Agent Profile: Banking apps now feature dedicated dashboards where users explicitly register their AI assistants (e.g., Google Assistant, Microsoft Copilot).34
- Granular Consent Flows: Users establish strict parameters, interacting with explicit prompts such as: "Google Assistant can spend up to $100 per day on groceries, but cannot make investments."34
- Cryptographic Identity: At the edge, agent wallets execute a four-step pipeline: hold funds, prove identity, check authority, and settle. To prove identity, agents utilize tokens (Skyfire's signed JWT specifying the model and capability scope) or headers (Visa's Trusted Agent Protocol, which signs the agent identity directly into HTTP requests using public-key cryptography).37
To facilitate this, infrastructure providers employ Protocol Compatibility Layers. This allows merchants to accept payments from various agent standards (ACP, AP2, MCP) without redesigning their checkout pages.38 For machine-to-machine transactions, rails like x402 utilize stablecoins to settle micro-transactions at a fraction of a cent on networks like Solana, bypassing traditional card interchange fees entirely.39 Agentic commerce fundamentally shifts the control point away from the visual checkout screen and toward the backend API layer that verifies the agent's cryptographic mandate.
[8] Strategic Outlook: Safety by Design [source]
Retrofitting security onto a completed product generates clunky, reactive friction. The alternative is "Safety by Design," a framework codified by the ISO 31700 standard, which mandates that Trust and Safety (T&S) teams integrate directly into the product development lifecycle from inception.40
When threat-modeling occurs at the architectural stage, financial institutions launch products equipped with targeted positive friction, adaptive rate limits, and contextual reporting tools. This federated accountability model—where product, engineering, and marketing teams share responsibility for the safety impact of their features—replaces siloed compliance departments.41 By embedding transparency, precise data governance, and intelligent behavioral friction natively into the application, proactive fraud prevention strengthens, rather than degrades, the modern banking experience.
References
[1] Federal Trade Commission. (2025). "New FTC Data Show a Big Jump in Reported Losses to Fraud to $12.5 Billion in 2024." Alloy 2025 State of Fraud Benchmark Report. 2: UK Finance. (2026). "Annual Fraud Report 2026." Fintech Futures. 3: UK Finance. (2025). "Confirmation of Payee: The journey so far." Pay.UK. 4: BioCatch. (2025). "UK Finance Half-Year Fraud Report 2025." BioCatch Blog. 5: White, J. (2025). "2025 U.S. Financial Protection Satisfaction Study." J.D. Power. 6: Grant Thornton. (2026). "Mitigating APP fraud through behavioural economics." Grant Thornton Insights. 7: Losada, C. (2025). "The Role of Behavioural Economics in Understanding and Countering Fraudulent Tactics." ResearchGate. 8: Open Banking Implementation Entity. (2022). "CoP & CRM Draft Standards." Standards Open Banking UK. 9: ThreatMark. (2026). "Smart Friction." ThreatMark Insights. 10: Webb, R. (2025). "Make a Payment." Russell Webb Design. 11: Sekura.id. (2025). "Fraud in 2022: five things you need to know." Malta CEOs. 12: Qonto. (2026). "How UX design can prevent scammers from stealing money." Qonto Way Medium. 13: ClearBank. (2026). "Confirmation of Payee API: How to implement CoP in your payment flows." ClearBank Insights. 14: Raman, A. (2025). "Pay.UK discusses the evolution of Confirmation of Payee with Tell.money." Pay.UK. 15: Eftsure. (2025). "Confirmation of Payee." Eftsure Finance Glossary. 16: Matsuda, T. (2025). "How Starling Bank turned a £29 million fine into an AI-driven strategy for trust." Medium. 17: Google Cloud. (2026). "Starling puts the power of AI in customers' hands with Gemini and Google Cloud." Google Cloud Customers. 18: Starling Bank. (2025). "Starling Bank launches UK-first AI tool to combat scams." Starling Bank News. 19: Security Brief. (2026). "Starling launches AI scam-checker for five million." Security Brief UK. 20: Capital One. (2026). "AirKey: Preventing fraud with the tap of a card." AirKey. 21: Fraud Fight Club. (2026). "The AirKey and Capital One Story." Fraud Fight Club Sessions. 22: Yahya, F. (2025). "Behavioral Biometrics: Revolutionizing Security with Human Nuances." Faisal Yahya Insights. 23: BioCatch. (2026). "A Top 5 UK Bank Prevents Account Opening Fraud and Saves £800K in First Three Months." BioCatch Resources. 24: BioCatch. (2026). "Top Asia Pacific Bank Rapidly Deploys BioCatch Account Opening Protection." BioCatch Case Studies. 25: Feedzai. (2026). "AI-Native Financial Crime Prevention." Landbase Blog. 26: Ademero. (2026). "Banking Fraud Detection AI." Ademero Resources. 27: Head of AI. (2026). "Monzo's AI-Driven Fraud Detection in Digital Banking." AI Industry Case Studies. 28: Bayton, J. (2026). "Reviewing the 2026 Android Security Paper." Bayton Blog. 29: Android News Wire. (2026). "June 2026 Android Drop: Fake Call Detection & 7 New Features." Android News Wire. 30: Aberdeen. (2019). "Modern Fraud Prevention Programs." NICE Insights. 31: J.D. Power. (2024). "Fraud Protection and Resolution Become Critical Components of Customer Experience." The Credit Union Connection. 32: DataDome. (2024). "Loyalty Fraud." DataDome Learning Center. 33: Rivo. (2026). "27 Fraud Detection in Loyalty Programs Statistics Every Ecommerce Brand Must Know in 2025." Rivo Blog. 34: G&CO. (2026). "Banking App Design Trends 2026." G&CO Insights. 35: Outcrowd. (2026). "Fintech Design Trends 2026." Outcrowd Blog. 36: Masood, A. (2026). "Agentic Payments 101." Medium. 37: Eco. (2026). "Agent Wallets: How AI Agents Spend Money." Eco Support. 38: Nuvei. (2026). "How Agentic Commerce is Redefining the Payment Landscape for Global Merchants." Nuvei Posts. 39: PayAI. (2026). "Agentic Payments." PayAI Network Blog. 40: Lallemand, C. (2024). "Ethical Tensions in UX Practice: Manipulation and Vulnerability." University of Luxembourg. 41: WNS. (2026). "The Mortgage Rabbit Hole: Can You Still Afford to Stay In?" WNS Perspectives [source]