Facephi presents The New Fraud Frontier 2027, an analysis of emerging threats, their impact across six regions and the risks organizations will need to anticipate

The technology company’s report finds that less than 1% of what is lost to fraud worldwide is ever recovered. The blind spot lies in the method: companies fail to detect their own attacks or trace where their losses come from, and the gap between adopting agentic AI and governing it keeps widening.

Alicante, Spain November 19, 2026., Nov. 19, 2026 (GLOBE NEWSWIRE) — In the digital world, the line between legitimate activity and fraud is getting harder to draw. Attacks no longer always come through a stolen identity or an impostor: they can rely on real customers manipulated in real time, on synthetic identities capable of passing onboarding checks, or on digital content injected before it ever reaches the camera. On top of that comes a new capability: artificial intelligence agents that can plan and carry out fraudulent operations at scale and around the clock.

This is the landscape examined in The New Fraud Frontier (Foresight 2027), the Facephi Observatory report built on official sources, industry analysis and the company’s own expertise across six regions. The study points to a worrying conclusion: many organizations still cannot identify their own attacks or the real source of their losses. The problem lies not only in how sophisticated the threats are, but also in the blind spots of the methods used to detect them.

The arrival of frontier AI 

Agentic AI can already plan and execute fraud campaigns end to end, and AI-powered fraud is estimated to be 4.5 times more profitable than traditional fraud, according to INTERPOL. This reinforces trends identified in previous years, such as Fraud as a Service and the integration of AI into criminal operations, which broaden access to increasingly sophisticated attack tools and techniques.

That is why frontier models are no longer treated as ordinary software: in 2026, governments began managing them as strategic assets, subject to export controls and “sovereign AI” strategies. Recent cybersecurity incidents and the growing number of vulnerabilities discovered with AI confirm this, according to the report. Meanwhile, adoption is moving faster than oversight: nearly 75% of organizations plan to use agentic AI within two years, but only 21% have a mature agent governance model.

The danger of dormant identities in bank accounts 

The fraud that causes the greatest losses no longer happens at onboarding, but afterwards, once the account is already trusted. The report describes the “dormant identity”: an account opened with flawless verification, sometimes using a synthetic identity, that sits inactive for months and then wakes up ready to receive a fraudulent payment, act as a mule or launder money.

“We have to stop asking customers whether they are who they say they are and start asking whether they are still who they said they were,” says Javier Barrachina, R&D Director at Facephi.

This fraud is rarely recorded as such. It surfaces weeks or months later as a default, account risk or a credit loss, among other things.

In other cases, there is no impostor. In authorized push payment (APP) fraud, the customer is real, uses their own device and approves the transaction themselves, persuaded by someone impersonating their bank, a public authority or a service provider.

What to expect in 2027: from account takeover to agent takeover 

Facephi anticipates two shifts. Fraud will spread beyond regulated industries: travel, gaming, e-commerce and the public sector will inherit problems that once seemed exclusive to banking, because attackers go where resistance is lowest. And a growing share of those acting in digital channels will not be people. AI agents are already starting to buy, pay and handle tasks on behalf of their users, and there are documented cases of agents that accessed systems without authorization while pursuing a legitimate goal. “Just as we talk about account takeover today, tomorrow we will be talking about agent takeover,” the report warns. Added to this is the debate over cryptographic obsolescence.

In 2027, preventing fraud won’t be enough: organizations will have to prove it. Across the six regions analyzed, regulation is moving beyond one-off identity checks toward continuous, evidence-based oversight.

Against this backdrop, Facephi proposes reading identity as a continuum in context: verification, session, account and transactions correlated on a single record.

“We have to stop asking customers whether they are who they say they are and start asking whether they are still who they said they were,” Barrachina sums up.

The New Fraud Frontier covers North America, Latin America, Europe, the Middle East and North Africa, Asia-Pacific and Africa, and includes a five-level maturity model, ten questions for the risk committee and seven indicators to measure progress. The full report is available here.

About Facephi 

Facephi is a technology company specializing in digital identity verification, fraud prevention, and regulatory compliance solutions. It delivers an AI/ML-powered Identity & Anti-Fraud platform that provides a comprehensive 360° end-to-end approach to digital protection. Renowned for its strong focus on security and data integrity, Facephi develops solutions that make digital processes safer, more accessible, and fraud-free. Its technologies help prevent identity theft while ensuring the ethical management and protection of personal data.

With over a decade of experience in developing technologies aimed at safeguarding digital identity, Facephi is headquartered in Spain, with subsidiaries in APAC, EMEA, LATAM and NORAM. The company serves the needs of clients across 30+ countries, delivering innovative solutions that address security challenges in an ever-evolving digital landscape.

Attachment

Irene Hernández Martín
Facephi
[email protected]

GlobeNewswire Distribution ID 9881203