Mercuryo warns AI crypto scams outpacing security

TLDR

Mercuryos Ashna Vaghela says AI driven crypto scams are evolving faster than current security controls, pushing more risk onto everyday users and eroding market trust.

1. Crypto scam losses surged to at least $14 billion in 2025 and are rising, with AI boosting the size and sophistication of attacks.
2. AI tools let fraudsters clone brands, build synthetic identities, and run deepfake outreach that bypasses traditional, code focused security.
3. The industry is shifting toward invisible risk checks, better UX, and clearer regulation, but users still need strict verification habits.

Deep Dive

1. Growing Scale Of AI Assisted Fraud

Mercuryos chief customer officer highlights that generative AI is helping scammers move faster than traditional fraud systems can adapt, especially in consumer payments and crypto rails. Chainalysis data cited in the interview estimates crypto scams received at least $14 billion in 2025, up from $9.9 billion in 2024, while average victim payments jumped about 253 percent, from $782 to $2,764 per case.AI driven crypto scam losses

These losses sit on top of broader fraud trends, with global scam reports also hitting record levels. In crypto, this contributes to a trust deficit that shows up in things like spot Bitcoin ETF outflows and heightened fear readings in sentiment indices when large hacks or scam waves hit.

2. Why AI Is Outpacing Existing Defenses

Vaghela stresses that the battlefield has shifted from software vulnerabilities to human manipulation. Generative AI makes it easy to:

1. Create synthetic identities and polished corporate style messages that look legitimate.
2. Clone brands and interfaces, including fake support chats or payment flows.
3. Produce real time deepfakes in voice or video that pressure victims to approve transactions.

Legacy fraud systems are mostly tuned to catch unusual card behavior or known malware, not subtle social engineering that rides on familiar apps and trusted logos.Mercuryos security critique

She argues the industry relies too heavily on education and complex UX, effectively blaming users for failures instead of redesigning infrastructure so the safest path is the default.

3. Emerging Defenses And Practical Checks

Vaghelas core proposal is more invisible security: automated micro checks inside transaction flows that verify identities, devices, and context without making users jump through obvious hoops. One example is Mercuryos work on Mastercards Crypto Credential, which replaces raw blockchain addresses with readable, verified aliases and supports biometric, card linked checkouts that can embed stronger risk controls.Alias based crypto credentialing

She also points to risk based, clear regulation as a positive, not a blocker, because it gives payment providers and crypto firms concrete standards for fraud monitoring and consumer protection. On the user side, practical habits remain critical: treat unsolicited investment or urgent security messages with suspicion, verify any upgrade or RWA token story through official project sites, and use strong account protection (hardware key style 2FA, unique passwords, strict app permissions).

Conclusion

AI is amplifying crypto scam capabilities faster than many security stacks are evolving, shifting attacks toward convincing, human targeted fraud rather than obvious malware or code exploits. That dynamic undermines user confidence and can feed into broader market stress when big incidents coincide with price drawdowns or ETF outflows. As infrastructure vendors roll out alias based credentials and embedded risk checks, the most resilient setups will be those that reduce user friction while quietly enforcing strong verification, leaving education as a supplement rather than the only line of defense.