AI fuels new-account fraud
- Reports show overall identity‑fraud losses stabilized in 2025, but new‑account fraud surged and AI-driven scams rose. - Analysts warn AI enables unsophisticated actors to scale sophisticated fraud tactics quickly. - Rising AI-enabled fraud creates new trust and prevention challenges for platforms, banks, and consumer services ( ).
Identity-fraud losses were mostly flat in 2025, but criminals opened more new accounts in other people’s names and used artificial intelligence to make scams harder to spot. (javelinstrategy.com) Javelin Strategy & Research said the combined impact of identity fraud and scams totaled $38 billion in 2025, down from $47 billion in 2024, while the number of victims fell by 4 million to 36 million. Identity-fraud losses excluding scams were $27.3 billion in 2025, essentially unchanged from $27.2 billion a year earlier. (javelinstrategy.com) The shift was inside the totals: new-account fraud rose, account-takeover fraud remained the costliest category, and scam losses fell 45% year over year to $10.7 billion from $19.5 billion. Javelin said some of that decline reflected more scams that did not cause immediate cash losses but still exposed victims to later account compromise. (biometricupdate.com) (markets.businessinsider.com) New-account fraud is the kind that starts with a fake application for a credit card, loan, bank account, or phone plan using stolen personal data. It hits lenders, fintech apps, telecom companies, and marketplaces because the fraud happens at sign-up, before a customer relationship is established. (javelinstrategy.com) Analysts say artificial intelligence lowers the skill needed to run those schemes. Tobias Thonak, a BearingPoint partner, told finews that AI lets “even the most foolish” fraudster produce more convincing messages, fake identities, and attacks at much larger scale. (finews.com) That changes the economics of fraud for banks and platforms. A criminal no longer needs to write clean phishing emails, build passable fake documents, or imitate a voice by hand when cheap tools can generate text, images, audio, and video in seconds. (aarp.org) (finews.com) Consumers are seeing that in everyday scams. AARP said easy-to-use AI tools are helping criminals create cloned voices, deepfake videos, and realistic messages for impostor schemes, pushing more organizations to tell users to verify requests through a second channel. (aarp.org) Javelin said the result is not just financial loss but weaker trust in digital services. When fake applications, impersonation calls, and synthetic identities become cheaper to produce, banks and consumer apps face more pressure to tighten onboarding checks without turning away legitimate customers. (javelinstrategy.com) The 2025 numbers suggest fraud is moving from obvious theft toward quieter account creation, data harvesting, and delayed misuse. Loss totals may look steadier on paper, but the fight is shifting to the moment a new customer first shows up. (javelinstrategy.com)
