The reality of financial crime is that the same neural networks that can generate the perfect fake voice or clone a customer’s face can receive training Detect those counterfeits faster than human analysts. This paradox lies at the heart of where Instant payments can easily become instant fraud.
“the very interesting The conundrum we find ourselves in.” Entersect CEO Schalk Norte he told PYMNTS. “AI democratizes attack vectors. You no longer need to be an expert to conduct an attack.”
But despite the noise about synthetic personal information and deepfakes, the principles underlying digital trust have not yet been established, Nolte said. rewritten.
“If you take a step back, the foundations of authentication and security have not been established. influenced” he said.
Nolte said that only by combining AI-driven insights with timeless security fundamentals and collaborating across institutions can banks stay ahead of a fraud ecosystem that learns at machine speed.
Advertisement: SCROLL TO CONTINUE
The unchanging core of authentication
What is clear is that the rise of generative AI has lowered the technological barrier to entry, allowing anyone with a laptop to launch phishing campaigns that once required specialized code and infrastructure. Accessibility speeds up attacks, and of defense, Variables that determine bank security.
Also, Defense multiplier. Through the consortium, banks can flag fraudulent devices, IPs, or transaction signatures before they cause more damage.
Nolte said the pattern of fraud repeats at different institutions.
“The attacker attacks this bank and then the next bank,” he said. “If you only look at your own data, you may not notice patterns.
“Typically in the first instance, people can get away with a crime unless they have intelligence,” he added. “But the second time, we can stop it. With a consortium, we can avoid the losses of the first time. … If we actually partner, we’ll have enough data to see patterns. ”
Equally important to technical defense is education, helping customers understand what banks will and will not require.
“Someone was calling customers pretending to be from their bank,” Nolte said, citing an example. “They just asked, ‘Have you seen any strange deals?'” The customer said “yes” or “no” and hung up. What the scammers did was record a voice and use AI to clone it and bypass the bank’s voice verification system. ”
This attack did not require stolen passwords or OTP codes, only human trust. Nolte also emphasized the need for authentication to be a dynamic process rather than a static checkpoint.
“You can’t rely on the fact that you sound like John on the other side,” he said. “You get two elements, you get two channels. … The same biometrics can also be used to test other signals. ”
The strategic future of AI in authentication
As AI becomes more deeply integrated into authentication, it is also reshaping how risk itself is measured.
“Free size model” It’s broken” Nolte said, adding that banks still rely on rigid playbooks such as push notifications and fallbacks to SMS, unintentionally inviting attackers to reverse engineer their defenses.
Instead, he said, banks should adopt dynamic orchestration and use AI to select the appropriate “treatment” for each risk, much like a doctor selects a drug based on a diagnosis.
We want to be your favorite news source.
Add us to your preferred sources list to see our news, data, and interviews in your feed. thank you!
“Antibiotics are effective if you have an infection, but they are less effective for headaches,” Nolte says.
Entersekt’s proprietary platform applies that logic in real-time, analyzing behavioral, device, and network data to determine which authentication methods to deploy.
“If this is a phishing attack, you put one mechanism in place, and if it’s a man-in-the-middle attack, you put another mechanism in place,” Nolte said. “Our philosophy from the beginning has been to strike a balance. [user experience (UX)] With security. ”
The challenge is integration. He said many financial institutions use multiple authentication vendors and their systems have “no input to each other.” Without a unified view, signals are siloed and fraud can slip through the cracks.
AI’s role is to connect those dots. Contexts turn static credentials into living profiles.
“We collect about 8.5 billion data points a year,” Nolte said, and this scale allows Entersect’s models to learn what “normal” looks like for each individual.
“The key is to determine what is normal for John, so you can determine what is abnormal,” he says.
The future of trust
As AI becomes embedded in financial structures, its role is shifting from reaction to prediction. The next generation of authentication will combine biometrics, behavior and consortium intelligence into a self-adjusting “trust structure” that recalibrates with each interaction, Nolte said.
He said he envisions a system that understands users by their rhythm rather than their credentials.
“I’ve seen you doing e-commerce on this device,” he said. “How can we use this to make banking safer and vice versa?”
Growing banks don’t just detect fraud faster. Intelligence can also be shared more quickly.
“This is still a big bullet, thick armor game,” Nolte said.
The future of trust belongs to those who build that armor together.
