Online payment fraud to exceed US$362 billion globally: Juniper

As online fraud continues to grow as e-commerce spending does, a new Juniper Research study highlights that online payment fraud will surpass US$362 billion globally over the next five years.

The report also said there will be losses of US$91 billion alone in 2028.

According to Juniper, a rise in e-commerce transactions in emerging markets is driving this growth.

As technology continues to evolve and become more complex, so too does the way bad actors behave online. Juniper noted that merchants are facing new threats, such as an increased use of AI for attacks.

Online payment fraud is where cybercriminals conduct false or illegal transactions online, using a number of different fraud strategies, such as phishing or account takeover.

Cara Malone, research author at Juniper Research said fraud detection and prevention providers must educate their clients on the importance of data sharing, in order for the highest accuracy within their solutions.

“This is increasingly important with the growing use of AI, as it utilises a variety of data to examine patterns within fraud, which is extremely advantageous in a space where fraudsters usually attack at scale, rather than attacking a specific customer,” she said.

Juniper Research released its latest competitor leaderboard for 2023 as part of the study.

The Juniper new competitor leaderboard ranked the top 21 fraud detection and prevention vendors, using criteria such as the relative size of their customer base, completeness of their solutions and their future business prospects.

The research found that the leading players scored well based on the breadth of their anti-fraud orchestration capabilities and their use of AI for analysing trends in fraudster behaviour.

In order to stay ahead of the competition, Juniper authors noted that vendors must utilise data collected throughout the whole e-commerce process to further develop their fraud detection and prevention solutions through training and advancing AI models.