A fraud team can correctly verify every identity and still lose money. A consumer can pass every authentication check and still authorize a payment to a criminal. That’s the paradox facing fraud, risk and identity leaders right now: The controls are working as designed and fraud is still getting through.
The reason is structural. For years, organizations treated trust as a single decision made at onboarding. Verify the identity, approve the account and rely on low-friction account authentication across the customer lifecycle. That model assumed fraud only attacked at the front door. Today it attacks everywhere: the phone call that precedes a transaction, the login that follows a credential theft, the account that's been open for months before anyone notices it was never real.
81%
of digital fraud occurs after account creation 1
AI has accelerated this shift. It lets criminal networks industrialize attacks that once required manual effort: building synthetic identities, generating convincing documents, running scam scripts at scale. Industrialized fraud doesn't respect organizational boundaries, and no single signal or control inside one company's walls can see the whole picture anymore.
The next phase of digital trust depends on connecting signals across the identity, customer lifecycle, enterprise and collaborative networks. Below are five predictions for where that shift takes the market — and what fraud, risk and identity leaders should do to get ahead of it.
1. Consumers protect themselves from digital risk with dedicated security apps
Consumers won't wait for a breach notification to find out they were targeted. Over the next several years, they'll increasingly rely on AI-powered consumer applications that assess whether a call, message, website or payment request can be trusted before they act, not after.
- Market signals: This shift is already underway. Mobile applications process caller ID, branded business identification and reputation, spam and scam labeling, and verified messaging in real time. Consumers are gaining the ability to question an interaction in the moment rather than discovering after the fact they were manipulated.
- How to prepare: Organizations should prepare by moving past identity monitoring and post-event remediation. Combining identity, communications, device, behavioral and network intelligence helps flag manipulation earlier and gives consumers a chance to recognize risk before they take action they can't undo.
2. The login will start to disappear
The password is becoming a relic. Passkeys, trusted devices, digital wallets and reusable identity credentials are changing how consumers prove who they are. Instead of typing a password or answering knowledge-based questions at every interaction, consumers will increasingly authenticate through persistent credentials tied to a trusted device, often with no visible login step at all.
- Market signals: Passkeys are already deployed across major technology ecosystems. Overall, 93% of accounts were eligible for passkeys while 36% had a passkey enrolled and only 26% of sign-ins used them.2 Digital identity wallets — including Apple Wallet, Google Wallet and the EU Digital Identity Wallet, along with national digital identities like Aadhaar in India and Singpass in Singapore — allow consumers to prove a specific attribute like age or eligibility without exposing their full identities. Authentication is becoming persistent and credentials are becoming selective.
- How to prepare: Organizations should design authentication strategies around continuous identity assurance, established across devices, credentials and interactions rather than a single identity verification moment. Fraud controls need to evaluate the device, behavior and context surrounding a credential — not just the credential itself. One caution applies here: Passkeys and wallets aren't yet standardized from an end-user experience perspective, leaving them open to spoofing. Backup and recovery pathways, where legacy authentication like one-time passcodes (OTPs) still gets used, remain the weakest link. Fraudsters will keep attacking it.
3. Fraud defense becomes a network — not a product
No organization can see the whole fraud picture by itself anymore. Scams and fraud schemes span institutions, industries, channels and digital ecosystems. A fraudster blocked by one bank's controls can easily move on to a telecom provider to a retailer to another bank using the same synthetic or stolen identity each time. As a result, fraud prevention is evolving toward collective defense, combining trusted information sharing and digital intelligence to reveal relationships and patterns no single organization can see on its own.
- Market signals: This shift already has real infrastructure behind it. Cifas in the UK,3 Stop Scams UK4 and the Anti-Scam Centre in Singapore5 all operate as collective defense models. In the US, the Financial Crimes Enforcement Network (FINCEN) issued updated guidance in June clarifying how financial institutions can share suspected fraud information under Section 314(b) of the USA PATRIOT Act, strengthening the legal foundation for collaborative detection.6
- How to prepare: Organizations should evaluate where consortium intelligence, privacy-preserving data collaboration and cross-industry partnerships can extend their own fraud defenses. Competitive advantage won't come from hoarding more data. It will come from connecting signals across the ecosystem fast enough to stop a threat before it becomes a loss.
4. Proving identity will no longer be enough to establish digital trust
Knowing someone is who they claim to be remains essential, but it's no longer sufficient. A legitimate consumer can be manipulated into authorizing a fraudulent payment. A legitimate account can fall under a fraudster's control. A legitimate AI agent could execute an instruction that doesn't reflect what its owner intended. The identity can be real while the interaction behind it isn't.
- Market signals: Scam and authorized fraud growth already demonstrate this gap. Identity and authentication controls can function exactly as designed while a consumer is simultaneously manipulated through a separate channel. Session hijacking, social engineering, deepfakes and AI-generated communications are making the line between identity and trust harder to see. As digital interactions increasingly involve humans, devices and AI agents acting on someone's behalf, understanding who’s acting matters, but understanding their intent and context is just as important.
- How to prepare: Organizations should enhance their fraud strategies to connect identity verification with behavioral, device, communications and network intelligence across the full customer lifecycle. The operating question needs to shift from "Is this the right person?" to "Should this interaction be trusted?"
5. Fraud intelligence will adopt a common operating language
Fraud teams across institutions describe the same attack patterns with different terminology, which slows collaboration precisely when speed matters most. As fraud becomes more interconnected, automated and global, the industry will move toward standardized taxonomies for describing fraud tactics, techniques and behaviors — the way cybersecurity built shared frameworks for describing adversary behavior.
- Market signals: The shift has already started. MITRE's Center for Threat-Informed Defense launched the Fight Fraud Framework,7 a behavior-based model that gives the industry a shared taxonomy for financial fraud tactics. Modeled on the principles behind MITRE ATT&CK, the framework describes fraud through observable adversary behaviors instead of disconnected, one-off events.
- How to prepare: Organizations should begin aligning internal fraud classifications with emerging industry frameworks and build data architectures that support interoperable, machine-readable fraud intelligence. A shared language for fraud enables faster threat sharing, stronger consortium intelligence and AI systems that learn from patterns observed well beyond any single organization's walls.
Establish trust continuously as a foundation of sustainable growth
Every prediction above points to the same underlying shift: Trust can't be established once. It must be assessed continuously — across identities, devices, communications and relationships — as conditions change.
Organizations treating fraud prevention as a series of isolated checkpoints will keep losing ground to attackers who move faster than any single team can track. Organizations that connect signals across the full customer lifecycle gain something more valuable than a lower fraud rate: the ability to make quicker, more confident decisions about who and what to trust.
Explore recent trends and business fraud survey insights in the H2 2026 Update: Top Fraud Trends Report, Fraud in the Age of Continuous Trust.
