미래의 위협으로부터 디지털 자산 보호

In partnership withLedger Securing digital assets against future threats AI-enabled fraud and the coming impact of quantum computing are redefining digital-asset security, putting pressure on owners and service providers to act now. Cryptocurrency thieves are getting creative. Taking advantage of the desire to learn more about crypto and banking on the digital assets’ reputation as a way to get rich quick, AI-generated video tutorials are touting ways of make money from crypto-trading arbitrage — purportedly teaching viewers how to create maximal extractable value from trades using smart contracts. Yet, they are often scams, a subtle social engineering approach known as mentorship pretexting. In one case, a fraudster shifted more than $900,000 from victims’ accounts to their own wallets. AI chatbots and large language models have allowed the scammers to be more convincing as well: In 2025, about 60% of all inflows into scammers’ crypto wallets came from AI-powered scams, according to cryptocurrency intelligence firm Chainalysis. Two major technological advances—AI and quantum computing—are the impetus for significant innovation across industries. Unfortunately, the cybercriminal ecosystem is no different. Cybercriminals' experimentation with AI, the threat quantum computing poses to encrypted data, and the rapid adoption of digitized value are resulting in massive changes, says Ian Rogers, chief experience officer at Ledger, a provider of secure signer platforms. “We have lived through the ‘once in humanity’ digitization of all information, and now we are living through the ‘once in humanity’ digitization of all value,” he says. “And I would say, we may all have a bit of whiplash from the internet, but you ain’t seen nothing yet.” The ubiquity of AI and continuing advances in quantum computing will transform the security landscape and change what companies and users need to safeguard their digital assets. Quantum computing poses challenges for the cryptocurrency ecosystem, especially for those areas not updated to use post-quantum cryptography, while AI lowers the barriers to creating synthetic identities and convincing fake information. “We have lived through the ‘once in humanity’ digitization of all information, and now we are living through the ‘once in humanity’ digitization of all value,” he says. “And I would say, we may all have a bit of whiplash from the internet, but you ain’t seen nothing yet.” The impact? Unless companies and digital-asset owners adopt more stringent security, they face more advanced threats and risks to their portfolios. Disruption, but when? As demonstrated by the mentorship scam, AI already poses a threat to technology users. A variety of other AI-augmented attacks have popped up as well. Attackers use AI code generators to produce variations on their tools, often successfully evading malware detectors and antivirus software. In one instance, a cybercrime group known as GreedyBear generated 150 wallet extensions for Firefox using AI code-generators. The malicious campaign stole more than $1 million from users. “As a user, it is very difficult to know if you are interacting with a human or with a bot,” he says. “How do you know that you are, today, interacting with me and that I’m a human? Because it’s already quite easy for AI to impersonate me.” Increasingly, AI is being used to masquerade as executives at companies or create synthetic identities for fraud. The attacks are often very convincing, even fooling tech-savvy victims, says Charles Guillemet, chief technology officer at Ledger. “As a user, it is very difficult to know if you are interacting with a human or with a bot,” he says. “How do you know that you are, today, interacting with me and that I’m a human? Because it’s already quite easy for AI to impersonate me.” The threat posed by quantum computing to encrypted data is real, but it’s still in a future state. For example, it’s likely a quantum computer capable of storing a million qubits is needed to break today’s commonly used public-key encryption. However, even with accelerated investment in research and development a practical quantum computer will only be deployable in the next decade or two. However, while practical quantum computing may not be here today, sensitive data needs to start being protected now. Far-sighted crypto thieves—not to mention nation-state threat actors—can collect high-value data today in the expectation that the data will remain valuable when it can be decrypted in a decade. The scheme, known as “harvest now, decrypt later, ” means that today’s most valuable data needs to use post-quantum encryption to protect against the future development of a practical quantum computer. “It’s not that easy to evaluate the threat,” says Guillemet. “However, the good news is that we have a solution to this threat.” The entire cryptocurrency ecosystem needs to adopt post-quantum cryptographic algorithms to protect asset owners from these future threats