The truth about the 2025 crypto hacker year: human errors far exceed code vulnerabilities

Last year was recorded as the darkest year in crypto history—record-breaking losses and frequent hacking incidents—but the reality is more complex than this label suggests. Despite the staggering numbers, the most destructive threats do not stem from smart contract code flaws but from the most overlooked vulnerability: humans.

According to data from on-chain analysis firm Chainalysis, approximately $17 billion in crypto assets were lost to scams and fraud in 2025, making it the second-largest threat after infrastructure hacks. Even more shocking is the growth rate: impersonation scams increased by 1400% year-over-year, and AI-driven schemes are 450% more profitable than traditional scams. These figures reflect a turning point—traditional hacking attacks are gradually taking a backseat, with human error and social engineering scams becoming the new main actors.

Scams Rapidly Surpassing Hacks as the New Threat

A major case exposed in early 2025 highlights the danger of this trend. According to blockchain researcher ZachXBT, an attacker successfully stole digital assets worth $282 million through social engineering, causing victims to lose 2.05 million Litecoin (current price approximately $58.31/LTC) and 1,459 Bitcoin (current price approximately $77.67K/BTC). The stolen funds were quickly exchanged for privacy coins Monero and transferred.

This case exemplifies a new characteristic of modern hackers: no longer requiring deep coding knowledge, a carefully crafted scam call, a convincing phishing email, or a fake identity are enough to yield huge returns. Chainalysis data shows that impersonation and AI-driven scams are increasingly targeting individual users rather than infrastructure, marking a fundamental shift in attack patterns.

Defense Dilemma: Low Adoption of Tools and Widespread Vulnerabilities Coexist

Paradoxically, on-chain security is actually improving in tandem. Immunefi, an on-chain security platform, CEO Mitchell Amador stated in an interview, “As code becomes harder to exploit, hackers are turning to new, more cunning methods.” This evolution is redefining the landscape of security threats.

However, defense efforts lag significantly. According to Amador’s assessment, over 90% of crypto projects still have critical and exploitable vulnerabilities—something that ideally should not be the case. More critically, the adoption rate of defense tools is extremely low: less than 1% of the industry deploy firewalls, and fewer than 10% use AI-driven detection tools. This creates a paradox—security solutions exist, but most participants are still relying on outdated defense postures against the new wave of hackers.

AI Reshaping the Attack-Defense Ecosystem, New Security Focus Emerges

2026 will be a watershed year as artificial intelligence fundamentally changes the security landscape. Amador pointed out, “Defenders will increasingly rely on AI monitoring and response systems that operate at machine speed, while attackers are using the same tools for vulnerability research and large-scale social engineering attacks.” The arms race is accelerating.

But a more dangerous threat lies in on-chain AI agents themselves. When autonomous decision-making systems begin managing on-chain operations, new attack surfaces emerge. “On-chain AI agents may surpass human operators in speed and capability,” Amador warned, “but if their access paths or control layers are compromised, these systems also possess unique vulnerabilities unlike those of humans.” In other words, a hacked AI agent could cause more catastrophic damage than any traditional hacker.

From Code Defense to Human Management, the Security Paradigm Is Shifting

Data and expert opinions point to a clear trend: the battlefield of crypto security is shifting. As on-chain code becomes harder to exploit, hackers are hunting humans; as artificial intelligence intervenes, new forms of intelligent attacks are emerging.

Amador summarized the core implication of this change: “Human factors are now the biggest weak link that on-chain security experts and Web3 participants must prioritize.” This means that future security will no longer rely solely on smart contract audits and code reinforcement but will focus more on user interface design, enterprise-level access controls, real-time monitoring systems, and fundamental security awareness education.

In other words, 2025 was recorded as the darkest year for hackers, but the lesson is clear: the most lethal vulnerabilities are not on the blockchain but in human decision-making. Defensive upgrades must move upward from code—toward processes, management, and people.