BioCatch unveils DeviceIQ to spot banking fraud pre-login

BioCatch has launched DeviceIQ, a device intelligence product for banks that assesses risk in the moments before a customer begins authentication in a mobile app or web session.

BioCatch frames the product as a response to shifts in digital fraud, including device spoofing, emulators, and tools that obscure device attributes. It also points to the growing use of automation and agentic AI that can interact with banking services and alter the signals banks typically use to assess a session.

DeviceIQ analyses indicators in the milliseconds between a user opening a banking app and starting the login process. The aim is to spot compromised devices early, allowing banks to block access or flag a session before an attacker attempts account takeover or initiates a scam payment.

Banks have used device recognition for years as part of authentication and fraud controls, while tightening privacy practices and, in some contexts, reducing reliance on persistent identifiers. Meanwhile, fraud groups have adopted methods that reuse the same physical devices while masking identifying signals, weakening conventional device reputation systems.

Pre-login signals

DeviceIQ focuses on device "health" and trust rather than only whether a bank has previously seen a handset or browser. BioCatch says it can detect jailbroken devices, missing core sensors, debugging tools, and unauthorised code that attempts to intercept, monitor, or modify a banking app's activity.

It is also intended to reduce disruption for legitimate customers when they change phones or reinstall an app. DeviceIQ establishes a persistent identity across web and mobile environments, according to BioCatch, and is designed to recognise legitimate device upgrades without repeatedly prompting users to re-validate a device.

BioCatch says DeviceIQ is delivered through its Connect platform, which combines behavioural, device, transactional, and application intelligence. A single software development kit links DeviceIQ to the platform, it adds.

Cross-bank context

Another component draws on BioCatch's wider network of customers and products. BioCatch says it can provide context on whether a device has previously been associated with mule activity, scams, or account takeover at another institution, rather than offering only a blocklist of devices.

The cross-bank approach reflects a broader shift in fraud operations towards sharing indicators of compromise and patterns of abuse. UK banks have increased investment in this area amid rising authorised push payment fraud and scams that combine social engineering with fast-moving digital account access.

The UK government has also highlighted tech-enabled fraud as a national priority. BioCatch's launch outreach refers to a £250 million fraud strategy and a focus on AI-driven scams. The company argues that pre-login device checks can strengthen controls before a customer reaches password entry or biometric steps.

Agentic AI layer

BioCatch is also introducing DeviceIQai, an additional layer it says can distinguish between human-led sessions, hybrid sessions, and agentic AI sessions. The goal, it says, is to separate genuine automation from fraudulent activity.

DeviceIQai also flags potential deepfakes, according to BioCatch. It identifies signals such as virtual camera use, pre-recorded audio, and the use of video or images that could be used to bypass identity checks.

BioCatch argues that these trends expand the attack surface beyond the login screen. "Many financial institutions today rely on risk signals scattered across a patchwork of device and risk tools from multiple vendors," said Ayelet Eliezer, Chief Product Officer at BioCatch.

"That fragmentation not only drives up costs, complexity, and maintenance but also reduces efficacy, efficiency, and scalability. DeviceIQ is built directly into the BioCatch Connect platform, enabling banks to evaluate all risk signals in one place, giving them the clarity they need to stop fraud earlier, reduce friction for genuine customers, and prepare for a rapidly approaching AI-driven future," said Eliezer.

Industry analysts have also focused on earlier-stage signals in fraud prevention as criminals move faster and distribute attacks across channels. "The fraud prevention perimeter has moved," said Sam Abadir, Research Director of Risk, Financial Crime, and Compliance at IDC.

"Institutions that rely solely on identity signals at login are missing an earlier and increasingly exploitable attack surface," Abadir said. "Device-level intelligence collected before authentication gives risk teams a more complete picture of session context, which matters more as agentic AI blurs the line between legitimate automation and account takeover."

BioCatch says more than 340 financial institutions use its products and that it analyses billions of user sessions each month. It says it collects thousands of anonymised data points from digital interactions, including behavioural signals and physical device attributes, and applies AI and machine-learning models to distinguish legitimate customers from criminals.

DeviceIQ is entering a market where banks already use a mix of device fingerprinting, behavioural biometrics, and risk engines from multiple vendors. BioCatch's proposition centres on combining device assessment with behavioural analysis and pushing risk evaluation earlier in the customer journey, including before a user begins authentication.