Callsign using Dynamic Interventions ML to understand if user is acting under coercion

Chris Stephens, solutions engineer, EMEA, Callsign.
Chris Stephens, solutions engineer, EMEA, Callsign.
2 years ago

Callsign announced its Dynamic Interventions capability controlled by its Orchestration Engine. This technology allows organizations to detect social engineering scams in real time and intervenes when fraud is detected, delivering contextually relevant and personalized messages protecting the user from becoming a victim of fraud.

The phenomenon of social engineering has emerged as a way for fraudsters to psychologically manipulate consumers into transferring money to accounts held by the fraudsters. Known in the industry as authorized push payment fraud it is hard for banks to detect as the authorized customer is making the payment.

‘Scams like authorized push payments are a digital fraud problem requiring a digital solution. With real-time and faster payment systems becoming the global norm, banks and financial service organizations need to look beyond conventional fraud detection and actively detect, intervene and protect. With APP, static fraud warning messages to customers have become ubiquitous and easy to ignore. Fraudsters anticipate static messages and coach users past those warnings, but Callsign’s Dynamic Interventions detects threats to consumers in real time, intervenes and alters digital journeys appropriately, introducing new controls or steps to protect customers,’ said Chris Stephens, solutions engineer, EMEA, Callsign.

In the last 18 months, APP fraud has grown exponentially. Recent research from industry analysts Forrester reveals that authorized push payments are considered a major problem by 66% of financial services and consumer banking organizations across the globe (APAC 66%, Middle East 67%, North America 56%, UK 72%). Organizations agree that using a combination of threat detection, dynamic fraud warnings and behavioral biometrics can help solve this issue.

With faster or real-time payments in place across the globe, it’s impossible to recover money once it has been sent, ramping up the costs of fraud for financial services organizations or consumers. Zelle, a peer-to-peer banking app in the US, recently hit the headlines due to APP fraud occurring on the platform. Available to over 100 million banking customers, it is estimated that $490 billion was sent by consumers and businesses in 2021 over the Zelle network and because transactions are instant and non-reversible, the platform has attracted fraudsters.

In Asia 40% of banks named social engineering as the number one fraud concern with regards to real-time payments and 25% of consumers reporting they had encountered social engineering scams.

In the UK, research from UK Finance reveals customers lost £583.2 million in 2021 after being manipulated into transferring money into accounts controlled by fraudsters. The UK’s Faster Payments system was the payment method for 97% of those losses.

Using machine learning, the Callsign platform understands recurring behavorial patterns of users when they make online payments and uses that knowledge to detect if the user is acting under coercion. Combined with threat and malware detection, Dynamic Interventions intervene the moment a customer might be in danger, delivering intelligent, contextual and timely fraud messages to the consumer or stopping payments altogether. Crucially, for genuine users performing recognised activity these messages won’t be presented. This ensures users will not get message fatigue.

The platform works in unison and identifies threats by asking these questions:

  • Is the session secure?
  • Is the user human? Is there any malware present on the user’s device or has their device been compromised by a bad actor?
  • Is this user authorized? Is the user allowed to make the transactions they wish to make?
  • Is the user being tricked? Are they performing any actions that are unusual, such as making a big transfer to a suspicious account? Does their behavior indicate that they are under some sort of duress or being coached?
  • How can we manage the risks and user experience? What user journeys do we need to orchestrate?

For example, a bad actor may attempt to deploy an impersonation attack against a genuine user, by claiming to be a bank’s representative and asking for a large payment. As the legitimate user attempts to initiate the payment, Callsign’s Orchestration Engine detects something is wrong by screening the genuine user’s device to ensure that a fraudster isn’t using remote access software to take over the account – or that malware was running, ensuring that the fraudster can’t fool the technology by turning it off at specific stages in the digital journey.

From here, the genuine user’s typing cadence and mouse movements are monitored – for example, they may be slower and more ponderous, indicating they are on the phone receiving directions. Before the payment is approved, Dynamic Interventions powered by Callsign’s Orchestration Engine asks the user about the payment they are making, prompting them with multiple contextual questions. Based on customer answers, threat detection and customer behavior, the technology establishes if a bad actor is directing them. The answers confirm to Callsign they are under attack, the Orchestration Engine would advise them it is likely that they are under the influence of a bad actor, and to not comply with their instructions, or would intervene to prevent payment taking place.

When contrasted with the current setup that requires users to read and consent to generic messaging in a challenging user interface, this approach is more targeted, user friendly, and engaging. Organizations can customize the interventions to suit their sector and their customers’ needs and can be deployed with just a few clicks of a mouse without any need for coding. Consequently, fraud teams can establish interventions to counter new attacks quickly and efficiently, ensuring that bad actors aren’t successful.

Regulation needs to play a key role when it comes to enforcing consumer protection. Forrester’s research revealed that almost half of the financial services and consumer banking industry see regulatory issues as a significant challenge when trying to detect and prevent scams such as APP fraud. The impact of regulatory compliance has been cited as most significant in both the US and the UK, whereas organizations in the Middle East and Asia Pacific said the cost of victim reimbursement has significant impact for their organizations.

With the increase in APP fraud and consumer losses, regulators around the world are acting. The Consumer Financial Protection Bureau (CFPB) in the US recently revised guidance on unauthorized electronic fund transfers (EFTs) and, the UK Government and Payment Services Regulator (PSR) intend to improve reimbursement of APP fraud scam victims. In Asia, while the Monetary Authority Singapore is still updating the framework for equitable shared losses through scams for consultation this year, it doesn’t yet make provision for APP fraud.

It’s clear a new approach to combatting online scams is needed, and the best result for organizations in all sectors is layering solutions, for example using a combination of threat detection, dynamic fraud interventions and behavioral biometrics, to ensure genuine users are protected.

The Orchestration Engine gives banks the capability to manage identity and access management technology in one place. A key feature of Callsign’s Orchestration Engine is its no-code functionality which means banks can control and adapt customer journeys without a code release. This includes configuring Dynamic Interventions.

Don't Miss

Chris Stephens, Solutions Engineering, Callsign.

Security worse in Metaverse than social media

Callsign has predicted 2023 will be another year of fraud with emerging
Saeed Ahmad, Managing Director, Middle East and North Africa, Callsign.

Behavioural biometrics versus physical biometrics

Financial institutions have had to transform their services and business models’ countless