Crypto Scams Soar to $14 Billion—AI Outpacing Security Measures
By John Nada·Jul 18, 2026·5 min read
Crypto scams hit $14 billion, driven by AI that outpaces security. Mercuryo's Vaghela warns traditional systems can't keep up.
Chainalysis has reported a staggering $14 billion lost to crypto scams in 2025, marking a significant 253% increase in victim payments. This figure is more than a statistic; it's a warning signal for the crypto industry, highlighting a rapid evolution in fraud tactics that is outpacing existing security protocols.
Ashna Vaghela, the Chief Customer Officer at Mercuryo, explained to CCN how AI-driven scams have become increasingly sophisticated, focusing on exploiting human behavior rather than targeting technical vulnerabilities. The use of generative AI to craft synthetic identities and deploy deepfakes at a rate faster than current security systems can detect is reshaping the threat landscape. This growing challenge presents a paradox, where advancements in AI technology complicate efforts in cybersecurity.
The implications of these developments are extensive. The decline in Bitcoin’s value, exacerbated by geopolitical tensions and the negative sentiment arising from these security breaches, emphasizes the urgent need for a reinforced security infrastructure. Traditional fraud detection systems, built around outdated models, are proving ineffective against the agile and adaptable nature of AI-driven fraud.
Vaghela describes a rapidly morphing threat environment where retail-focused fraud is not only prevalent but also increasingly elusive. Legacy systems are struggling to keep up with the high-speed, adaptable tactics employed by AI-powered fraudsters, thereby putting financial infrastructures at risk of failing without immediate detection.
The evolution of AI-powered scams has taken a tactical shift over the past year, focusing more on personalized attacks that manipulate consumers into authorizing transactions themselves. This new approach diverges from conventional cyberattacks, which typically exploit software vulnerabilities. Instead, these scams leverage behavioral vulnerabilities, allowing criminals to modify their tactics faster than companies can update their static security systems.
"These human-centric threats evolve faster than software patch cycles because they exploit behavioral vulnerabilities rather than code bugs," Vaghela noted. This definitive shift underscores the growing prevalence of retail-focused fraud, making it both common and challenging to identify. The risk is particularly pronounced for payment providers relying on older fraud-detection systems, which may fail to detect synthetic identities and the real-time threats they pose.
Vaghela expressed concern over the "hyper-velocity and mutability of these adversarial tactics," warning that financial infrastructure providers could silently fail if they continue to depend on legacy fraud tools. These traditional systems lack the capability to detect real-time threats posed by synthetic identities, a significant vulnerability in the current security landscape.
To counter these growing threats while maintaining the user-friendly nature of crypto payments, Vaghela advocated for a shift towards invisible security measures. This would involve embedding risk controls directly into transaction systems, enabling the detection and prevention of synthetic identities without disrupting the seamless user experience that crypto users have come to expect.
China's $125 Billion Trade Surplus Can't Mask Domestic Stagnation
China's $125.
"By running these automated micro-checks invisibly, we can block deepfakes or synthetic identities at the fiat-to-crypto gateway without disrupting the one-tap payment experience users expect," Vaghela explained. This approach aims to maintain the balance between security and usability, which is critical for the growth and trust in the crypto ecosystem.
The challenge posed by AI-driven scams is not just a technological one but also a strategic one. The crypto industry needs to innovate beyond legacy solutions and adapt to a world where AI serves as both a tool and a threat. This dual nature of AI in the crypto space requires an adaptive and forward-thinking approach to security.
As AI continues to evolve, so too do the tactics employed by fraudsters. The ability to create synthetic identities, clone trusted brands, and produce real-time deepfake communications has added a new layer of complexity to the existing security challenges. This evolution necessitates a reevaluation of current security strategies and the implementation of more dynamic, AI-compatible solutions.
The impact of these developments is felt across the crypto industry. The negative sentiment fueled by scams and infrastructure hacks has contributed to Bitcoin's downturn, which is further aggravated by geopolitical tensions. Despite sustained institutional investor demand, the uncertainty surrounding Bitcoin and other risk assets continues to pose significant challenges.
In light of these challenges, the role of AI in both advancing and complicating security measures cannot be understated. The use of AI for malicious purposes highlights the need for equally sophisticated security solutions to protect users and maintain trust in the crypto ecosystem.
Vaghela's insights into the rapidly changing threat landscape provide a crucial perspective on the future of crypto security. As the industry grapples with these challenges, the need for innovation and adaptation becomes increasingly apparent. The integration of invisible security measures and the development of AI-compatible solutions are essential steps in addressing the evolving threats posed by AI-driven scams.
The crypto industry stands at a crossroads, where the choices made today will shape the future of security and trust. The rapid pace of AI evolution demands a proactive approach to security, one that anticipates future threats and adapts to the ever-changing landscape. As the industry navigates these challenges, the insights provided by experts like Vaghela will be invaluable in guiding the way forward.