How tech can win the AI arms race against fraudsters

Last year, significant events highlighted the rising threat of cyberattacks. Among them was the crush of phishing attacks corresponding with the launch of ChatGPT. According to SlashNet, there was a 1,265% surge between Q4 2022, when the groundbreaking generative AI tool was released, and Q4 2023. Scammers use GenAI to write more convincing phishing emails at scale. This is the latest escalation in the way cybercriminals seek to penetrate public and private enterprises to steal resources.

This costly trend underscores the need for the technology sector to work together to help organizations everywhere fight AI with AI. We need to collaborate to build and support platforms — often APIs that run quietly in the background — that allow AI to overlay multi-factor authentication processes. These processes add a step or two before logins or transactions to guard against stolen or hacked passwords, the phishing of personally identifiable information, and other attempted infiltrations. There are three essential steps to accomplish this goal.

First, AI should guide people to the most-secure login, and through their preferred communication channel. The dreaded password is finally fading, and “passwordless” solutions, powered by biometrics and other technologies, are on the rise. Authentication platforms should guide end-users to these options whenever possible. However, not every end-user will immediately adopt these new approaches. For the large majority, who are still authenticating through one-time-passwords (OTPs), AI can orchestrate the most-advanced, most secure way to get one. The goal is to leverage AI as the autopilot that zooms the passenger through the safest route, with the right amount of friction, balancing risk and access.

Once AI assists an end-user find the most secure log-in, it should also guide the interaction to the best channel that balances that risk with an optimal customer experience. The communication channel is driven by factors including consumer preferences, cost, time, place, and device. To deliver the best experience, allow for a range of channels, including SMS, WhatsApp, email, Viber, voice, and more. I am particularly excited about Rich Communication Services (RCS), which Google helped popularize and are coming to iOS this year. RCS messages can only be read by the sender and receiver, and branded with rich imagery, including a “verified” check mark. Alongside WhatsApp, Viber, and other channels, RCS has the functionality to reduce the number consumers who fall victim from social engineering and other communication scams. AI can play a valuable role by determining the secure channel a customer prefers and engaging them accordingly.

Second, use machine learning to outsmart bad guys. Machine learning can access historical indicators to uncover insights and identify causal relationships between unrelated variables indicative of unusual patterns that may precede fraud. It can continuously learn the latest threats by searching the dark web for emails, phone numbers, and other data exposed in data breaches. It can identify a swath of phone numbers or IP addresses being used for attempted intrusions. And it can use what it learns to assess the fraud risk of phone numbers, email addresses, IP addresses, and more.

AI can also detect anomalies based on a device’s location. Imagine a customer’s device sent signals 30 minutes ago that it was in San Diego, but now the same device looks like it’s in Syndney, Australia. AI can flag that instance as higher risk.

Third, AI determines friction and communication channels. AI should be programmed to dial friction up or down based on risk. Of course, friction — for example asking an additional security question — can have negative connotations. But when it comes to digital identity, it should be seen as a shield of protection to ensure safe transactions. Our research at Telesign shows that consumers overwhelmingly believe it is the companies they transact with that are responsible for their digital safety. Many have had their personal data stolen or an account infiltrated, which can be traumatic and time-consuming to repair. Through this lens, the right amount of friction should be welcomed, if communicated properly as a protective measure.

AI will help defend customers, and bottom lines, in 2024 and beyond. In July of 2023, the Securities and Exchange Commission instituted new rules requiring regulated entities to “disclose material cybersecurity incidents they experience.” It was a chilling moment that underscored how serious cybercrime has become. In the months that followed, two large companies disclosed attacks that disrupted sales, distribution, and revenue. Expect more such disclosures this year.

A planned defense needs to be part of the cost and planning for a modern customer experience model to protect businesses against the fraud and reputational risks that come these cyber-security incidents. AI, leveraged to inform the authentication process, should be the standard. Let’s build these platforms, and work together on interoperability, to make the digital world safer for everyone.

Written by Christophe Van de Weyer.

Have you read?
Top 50 global artists list, 2023: These are the most successful living artists in the world.
These Are The Most Popular Songs For Workplace Productivity.
Twenty-five of the world’s oldest billionaires in 2023 ranked.
Ranked: Countries That Produce the Most Carbon Dioxide (CO2), 2023.
Ranked: These Are The Most Desirable Jobs in the World, 2023.
Ranked: U.S. states by the number of billionaires, 2023.

This report/news/ranking/statistics has been prepared only for general guidance on matters of interest and does not constitute professional advice. You should not act upon the information contained in this publication without obtaining specific professional advice. No representation or warranty (express or implied) is given as to the accuracy or completeness of the information contained in this publication, and, to the extent permitted by law, CEOWORLD magazine does not accept or assume any liability, responsibility or duty of care for any consequences of you or anyone else acting, or refraining to act, in reliance on the information contained in this publication or for any decision based on it.