Zcash Co-Founder Warns AI Could Speed Up Cyberattacks on Crypto Before Quantum Does

The crypto industry has spent years worrying about a threat from the distant future: quantum computers powerful enough to break the encryption that protects digital assets. But a warning from one of the field’s most respected cryptographers suggests the more urgent danger may arrive much sooner, and from a different direction entirely. Eli Ben-Sasson, co-founder of Zcash and a computer scientist, warned in a post on X on November 5, 2026, that increasingly powerful artificial intelligence could accelerate cyberattacks against crypto wallets and the cryptographic infrastructure that underpins them. In his view, AI-driven security threats could become a more immediate concern than quantum computing, because AI is already here, improving fast, and being used by attackers today.

His message on X was blunt. “This quick rise in cybersecurity attacks is not going to slow down, not with AI becoming so powerful,” he wrote. “Your crypto keys, your online life, any app that holds something important to you — it’s all secured by cryptography.” The post, shared on November 5, 2026, went on to suggest that AI could help attackers well before quantum computers arrive, a framing that reorders the industry’s sense of which threat deserves attention first. Quantum computing remains a real long-term challenge, but it is a challenge measured in years. AI-assisted attacks are a challenge measured in weeks and months, and they are already unfolding.

AI Cyberattacks Crypto Security: Why AI Changes the Attacker’s Advantage

To understand Ben-Sasson’s warning, it helps to think about what AI actually does for an attacker. Traditionally, finding a vulnerability in software has been slow, skilled, manual work. A security researcher reads code, tests hypotheses, and probes for weaknesses one at a time. AI systems can now automate much of that process, scanning vast amounts of code, spotting patterns humans would miss, and generating attack attempts at a speed and scale no human team can match. The result is not a new kind of attack so much as a dramatic acceleration of the old ones.

Ben-Sasson argued that AI is changing the capabilities available to attackers, potentially allowing them to identify vulnerabilities faster than traditional cybersecurity methods. That speed matters enormously in crypto, where a single exploited weakness can drain millions of dollars in minutes and where transactions, once confirmed, are extremely difficult to reverse. An attacker who finds a flaw first wins, and AI is shortening the time between a flaw existing and someone finding it. Defenders must now patch faster, audit more thoroughly, and assume that any weakness will be discovered sooner than it would have been a few years ago.

The scale of AI assistance is equally important. A single attacker with a powerful AI system can run the equivalent of a large team of analysts, probing many targets at once. Phishing campaigns can be personalised at scale, malicious software can be adapted automatically to evade detection, and social engineering messages can be written in fluent, convincing language for any target. None of these techniques is new, but AI makes each of them cheaper, faster, and harder to distinguish from legitimate communication. For crypto users, that means the familiar threats, phishing links, fake support messages, and malicious downloads, are becoming more convincing and more common.

Zcash Eli Ben-Sasson Warning: Why Quantum Is the Wrong Priority Right Now

Much of the crypto industry’s security planning has been organised around quantum computing. The logic is straightforward: today’s widely used cryptography relies on mathematical problems that are hard for ordinary computers but could become easy for a sufficiently powerful quantum machine. If such a machine arrived, it could in theory break the encryption protecting private keys and, with them, the funds those keys control. That prospect has driven years of research into quantum-resistant cryptography and migration plans for major blockchains.

Ben-Sasson’s argument does not dismiss the quantum threat. It reorders it. He suggested that AI-powered attacks could develop into a serious security concern before quantum computers become capable of breaking widely used cryptographic standards. Quantum machines that can break real-world encryption do not yet exist, and building them remains an enormous engineering challenge. AI systems capable of accelerating attacks, by contrast, already exist and are improving rapidly. The threat that is already here deserves at least as much attention as the threat that may arrive in a decade.

This distinction matters because cryptocurrency security depends on multiple layers, including wallet software, private-key management, smart contracts, and network infrastructure. Even when the underlying cryptographic algorithms remain mathematically secure, weaknesses in how those algorithms are implemented can expose users to financial losses. A wallet with a flaw in its code, a smart contract with a logic error, or a key-management system with a weak point can all be exploited without anyone breaking the cryptography itself. AI is particularly good at finding exactly these kinds of implementation flaws, because it can examine enormous amounts of code and test countless variations far faster than a human auditor.

ZK-STARKs Quantum Resistance: A Hash-Based Path to Stronger Security

As part of his warning, Ben-Sasson pointed toward hash-based cryptographic techniques, particularly ZK-STARKs, as relevant technologies for building more resilient systems. ZK-STARKs, short for zero-knowledge scalable transparent arguments of knowledge, are a type of proof that allows one party to demonstrate that a computation is valid without revealing the underlying information. That capability supports privacy and verification at the same time, letting users prove something is true without disclosing the sensitive details behind it.

ZK-STARKs also offer an important technical advantage because their security relies on hash functions rather than the elliptic-curve assumptions commonly used in other cryptographic systems. Hash functions are among the oldest and most studied tools in cryptography, and they are generally considered more resistant to quantum attacks than elliptic-curve methods. That makes ZK-STARKs particularly relevant to discussions about long-term resistance to quantum attacks, and it is one reason Ben-Sasson, whose research helped develop the technology, has long advocated for their use.

The broader lesson is that crypto security needs to advance on several fronts at once. Quantum-resistant cryptography, including hash-based approaches like ZK-STARKs, addresses the long-term threat. Better code auditing, much of it AI-assisted, addresses the near-term threat of faster vulnerability discovery. And better user education addresses the oldest threat of all: attackers who bypass technology entirely by deceiving people. Ben-Sasson’s warning connects all three, because AI is now accelerating the second and the third while the industry’s attention has been fixed on the first.

What AI-Accelerated Attacks Mean for Wallets, Exchanges, and Everyday Users

The practical consequences of Ben-Sasson’s warning land differently depending on who you are in the crypto ecosystem, but nobody is untouched. For exchanges and wallet providers, the message is that security budgets and practices built for a slower threat landscape may no longer be enough. Code that was audited once a year may need continuous review. Vulnerability disclosure programs may need to move faster. And incident response plans need to assume that attacks will arrive more quickly and in greater volume than before.

For developers building smart contracts and decentralised applications, the warning is a reminder that audits are necessary but not sufficient. A contract that passed an audit six months ago may contain a flaw that today’s AI tools can find in hours, and the cost of that discovery falling into the wrong hands is measured in drained funds. Continuous testing, formal verification where possible, and bug bounty programs that reward researchers for finding flaws before attackers do are becoming standard practice rather than optional extras.

For everyday users, the implications are more personal and more immediate. Private keys remain the single most valuable target, and AI-assisted phishing is making the old advice, check the sender, look for typos, be suspicious of urgency, harder to follow, because the messages are getting better. The practical defences have not changed, but they matter more: use hardware wallets for significant holdings, keep seed phrases offline and never share them, enable every available security feature on exchanges, and treat any unexpected message asking for access or action as suspicious by default. No tool can replace that caution, and AI is making it more necessary, not less.

FAQ

Who is Eli Ben-Sasson and what did he warn about?

Eli Ben-Sasson is a computer scientist and co-founder of Zcash. In an X post on November 5, 2026, he warned that increasingly powerful AI could accelerate cyberattacks against crypto wallets and cryptographic infrastructure, becoming a more immediate concern than quantum computing.

Why could AI be a bigger near-term threat than quantum computing?

Quantum computers capable of breaking widely used encryption do not yet exist, while AI systems that speed up vulnerability discovery already do. Ben-Sasson argued AI is improving the speed and scale of attacks right now, whereas the quantum threat remains years away.

What is protected by cryptography beyond cryptocurrency?

Cryptography protects private keys, financial applications, personal information, and online accounts, not only crypto holdings. Ben-Sasson stressed that the same mathematical systems secure much of everyday digital life.

What are ZK-STARKs and why did Ben-Sasson mention them?

ZK-STARKs are zero-knowledge proofs that let one party show a computation is valid without revealing the underlying information. Their security relies on hash functions rather than elliptic-curve assumptions, making them relevant to long-term quantum resistance.

Do ZK-STARKs stop AI-assisted attacks?

Not on their own. They strengthen cryptographic foundations, but they do not automatically prevent AI-assisted software exploits, such as bugs in wallet code or phishing that tricks users into revealing keys. Defence needs several layers at once.

Source: KuCoin

Related reading on MM Crypto Analysis Web

• Blockchain Is the ‘Perfect Solution’ for AI Agent Trust, VeChain CEO Says
• AI Agents Raised Their Own Spending Limits in a Test — and Sent Money to an Impersonator
• Coinbase’s x402 Passes $42 Million as AI Agents Do 97% of Agent Payments

More in AI Crypto on MM Crypto Analysis Web.