Machine Learning and AI ‘Crucial’ to Fighting Fraud, Research Shows
Online payment fraud could cost companies more than $200 billion over four years, finds Juniper Research
- Written by Banking Exchange staff
Companies reliant on online payments stand to lose more than $200 billion to fraud over the next four years, according to new research.
Globally, companies in sectors such as ecommerce, airline ticketing, money transfer and banking services will all be hit as fraudsters use increasingly innovative scams, according to UK-based Juniper Research.
To combat the growing threat, organizations must take an “omnichannel” approach using cybersecurity, machine learning and artificial intelligence technologies to identify behavioral patterns in fraud attempts, the research firm said.
“The rapidly evolving nature of payment fraud and increased sophistication in attack methods requires machine learning adoption at scale, in order to minimize risk,” said Nick Maynard, co-author of the research. “Constant innovation in analytics and data models is increasingly essential to constraining fraudulent behaviors in payments.”
Juniper’s research also reported that machine learning tools could help payments industry stakeholders to “analyze transaction flows in a holistic way, unlocking hidden insights”.
It estimated that the incorporation of machine learning into fraud detection and prevention software would drive spending to reach $10 billion by 2024, marking a 15% increase on 2020.
“The increased convenience of digital solutions is driving greater ecommerce engagement,” Juniper’s report said. “However, it has also created a playground for cybercriminals intent on circumventing the structures that on which online payments rely. Understanding the threat landscape is crucial to reinforcing protections, whilst keeping innovation clear of exploitation by cybercriminals.”
Losses from digital money transfer frauds are expected to grow by 130% by 2024, Juniper stated, with emerging markets particularly vulnerable.
The report recommended that companies engage in ongoing ‘know your customer’ verification exercises, including “events-based re-verification”, in order to secure digital transactions.
Open Banking Issues
Juniper’s report – Fighting Online Payment Fraud in 2020 – also highlighted security concerns linked to the rise of Open Banking, the network of payment institutions and services facilitated by mobile apps and online platforms.
The US is among several major countries exploring the options offered by using the financial integration methods associated with the Open Banking ethos. Regulations have been introduced or are expected soon in regions such as Australia, the EU and Singapore to require financial services institutions to supply data to these services.
The research authors wrote: “Having open access to bank data, under user control and consent, is regarded by many countries as highly innovative in an era of hyperconnected ecosystems built on data.”
However, the authors warned that cybercriminals would “take full advantage of any delays” in implementing new regulations. For example, fraudsters had already engaged in phishing scams based on delays to the EU’s Secure Customer Authentication requirement, they said.
Tagged under Risk Management, Technology, AML & Fraud, Cyberfraud/ID Theft, Compliance/Regulatory, Operational Risk, Big Data, Security, Online, Cards, BSA/AML, Feature3, Feature, Financial Research, Payments,
- The federal response to recent banking developments — what is it and why does it matter?
- Regional Bank ETF Reflects Cautious Optimism
- Why Banks Must Focus on Personalized Experiences in Times of Economic Uncertainty
- March Appointments: New CFOs at Ally, Barclays
- Tech to Boost Community Bank Growth, Research Shows