Notifications
Clear all
Topic starter
11/07/2026 8:31 pm
TL;DR: Fraud operations now resemble SaaS businesses, with modular kits, on-demand botnets, and AI-enhanced social engineering compressing the time from compromise to extraction by 30% in 2024, according to AU10TIX. Static verification and batch controls are no longer enough when attackers can scale, adapt, and monetise within minutes.
NHIMG editorial — based on content published by AU10TIX: Fraud is morphing faster than ever in 2024-2025
By the numbers:
- 2024 saw a 30% reduction in dwell time between account compromise and monetary extraction.
- Synthetic identity fraud grew by 45% in 2024.
- Businesses report a 200% surge in attempted deepfake-aided wire fraud in Q1 2025 alone.
Questions worth separating out
Q: How should security teams respond to fraud-as-a-service and AI-driven identity abuse?
A: Security teams should treat fraud-as-a-service as an identity and access problem with financial impact, not as a separate fraud-only workflow.
Q: Why do deepfakes make identity verification less reliable?
A: Deepfakes weaken the reliability of voice, video, and conversational cues by making synthetic interactions look and sound legitimate.
Q: When should organisations prioritise real-time fraud monitoring over batch reviews?
A: Organisations should prioritise real-time fraud monitoring whenever compromise can quickly become monetary loss, such as account takeover, instant payments, or onboarding abuse.
Practitioner guidance
- Shorten the identity trust window Rework onboarding, recovery, and step-up decisions so they are tied to immediate behavioural and contextual signals rather than static checks that can be reused later in the session.
- Add synthetic-content resistance to proofing Test voice, email, document, and video verification paths against deepfake inputs and replay-style attacks, then require stronger assurance when the channel can be convincingly imitated.
- Move fraud containment into streaming controls Trigger quarantines, velocity caps, and manual review inside real-time monitoring pipelines so suspicious transactions can be paused before settlement or mule transfer completes.
What's in the full article
AU10TIX's full guide covers the operational detail this post intentionally leaves for the source:
- Step-by-step fraud detection patterns for synthetic identity, deepfake, and credential-stuffing scenarios.
- Operational examples of AI-driven monitoring signals and how they are tuned in live onboarding flows.
- Practical defence patterns for real-time transaction quarantine, step-up authentication, and remediation.
- The article’s own view of how KYC and AML requirements are changing across regions.
👉 Read AU10TIX's guide to AI-driven fraud trends and identity defences →
Fraud-as-a-service and deepfakes: what identity teams need now?
Explore further
11/07/2026 9:00 pm
Fraud has become an identity governance problem, not just a detection problem. The article’s central fact pattern is that attackers now operationalise onboarding, account recovery, and payment abuse as repeatable workflows. That means identity controls are being judged on whether they can preserve trust after first contact, not only whether they can verify at the door. Fraud teams and IAM teams therefore need a shared view of identity proofing, session trust, and recovery authority.
A few things that frame the scale:
- 80% of identity breaches involved compromised non-human identities such as service accounts and API keys, according to Ultimate Guide to NHIs.
- Only 5.7% of organisations have full visibility into their service accounts, which leaves large portions of machine identity estate outside routine governance.
A question worth separating out:
Q: What do identity teams get wrong about synthetic identity fraud?
A: The common mistake is treating synthetic identity fraud as a document problem alone. In reality, it is a lifecycle problem that combines fabricated identity attributes, weak verification, and delayed detection, so teams need controls that span proofing, behavioural signals, and ongoing transaction monitoring.
👉 Read our full editorial: AI-driven fraud is industrialising identity risk in 2025
