Article by: Oluwapelumi Adejumo
Compiled by: Chopper, Foresight News
Ethereum co-founder Vitalik Buterin publicly refuted a claim that artificial intelligence is predicted to cause Bitcoin’s price to drop more than 50% within two years.
The debate originated from the assessment by Silicon Valley investor and AI risk commentator Liron Shapiro, who believes there is a 50% chance that Bitcoin will drop more than 50% over the next two years, arguing that artificial intelligence will undermine investors’ trust in the security and robustness of the Bitcoin network.
But Vitalik holds the opposite view, posting on X: "My core logic is that I remain optimistic about cybersecurity in the long term; the real challenge lies in achieving the transition."
Vitalik suggested that Bitcoin's ecosystem can fully handle security issues that do not require societal consensus across the entire network, including attacks targeting clients, mining pools, and other network infrastructure. He believes AI is nearly impossible to break Bitcoin’s underlying hash function and proof-of-work mechanism.
He said he was originally willing to bet against Shapiro on this viewpoint, and that his existing crypto assets already represent about 90% of his total assets. He also believes this logic applies equally to Ethereum.
This divergence reflects a growing market split: what kind of changes will continuously evolving AI systems bring to the crypto market?
Shapiro’s argument does not require AI to break the underlying cryptographic algorithm of Bitcoin. Even if developers eventually patch the vulnerability, a large-scale attack that exposes security weaknesses investors believed to be extremely difficult to exploit would be enough to undermine market confidence.
AI attacks have already begun testing cryptocurrency defense systems
Real-world examples from the cryptocurrency industry demonstrate that the threat Shapiro warned about has arrived: hackers can create risks without breaking Bitcoin’s underlying cryptography.
In August this year, Boltz indefinitely suspended operations after suffering months of automated AI scanning attacks. Multiple controlled security breaches occurred in succession, and the small development team could no longer patch vulnerabilities as quickly as the attacks evolved.
Boltz stated that its non-custodial architecture ensures user fund security, but the company bore the losses from the attack, ultimately concluding that the business could not operate safely and sustainably.
This incident demonstrates that AI can pressure Bitcoin’s surrounding infrastructure without compromising the proof-of-work mechanism or breaking the underlying cryptographic system. Attackers rely on automated tools to systematically exploit vulnerabilities, operating at a speed far exceeding the defensive team’s ability to identify, patch, and deploy fixes.
Security industry executives expect these threats to spread widely as aggressive AI capabilities continue to escalate.
Deddy David, CEO of blockchain security firm Cyvers, previously told CryptoSlate that AI-driven smart contract attacks could ultimately result in economic losses of hundreds of millions to billions of dollars.
“If AI can大规模 exploit vulnerabilities in the core infrastructure of the internet, the crypto market will be the first to feel the impact,” said David.
Wallets, cross-chain bridges, exchanges, smart contracts, and the software that powers blockchain networks are all at risk. Compared to Bitcoin’s own consensus mechanism, these systems provide numerous entry points for hackers.
This is also the key point of divergence between Vitalik and Shapiro: AI does not need to break the SHA-256 hashing algorithm to cause serious harm to the entire Bitcoin economic ecosystem.
The central unresolved question is whether these attacks can be contained through upgrades and stronger defenses, as Vitalik expects, or whether they will ultimately severely undermine the security assumptions investors rely on when pricing Bitcoin.
Cryptocurrency and tech giants race to build AI defense capabilities
The entire tech industry is engaged in an arms race to ensure that defensive AI evolves at least as quickly as offensive AI.
Anthropic has restricted public access to its model because the model can autonomously discover software vulnerabilities and convert them into attack tools. Companies are instead delivering this technical capability to defensive security teams through the Glasswing program, with participation from technology and financial institutions including Amazon Web Services, Google, Microsoft, and JPMorgan Chase.
This project leverages Mythos Preview to identify and patch critical system vulnerabilities before malicious actors can exploit similar AI tools to launch attacks. Anthropic has committed up to $100 million in compute credits to support this initiative.
Participants in defensive strategies are no longer limited to individual companies. Over 100 organizations, including Google, Microsoft, Anthropic, and OpenAI, have jointly signed an open letter warning governments and businesses worldwide that AI-powered cyberattacks will become more common and sophisticated within just a few months.
Banks, payment giants, and tech companies such as Capital One, Mastercard, Visa, Adobe, Oracle, and IBM have also joined the joint appeal. They argue that as AI capabilities advance, existing cybersecurity defenses will no longer be sufficient, leaving critical infrastructure—already resource-constrained—more vulnerable. They urge governments and tech companies to provide advanced defensive AI tools and security testing services to institutions such as hospitals and water utilities.
The cryptocurrency industry is also caught up in this race and faces additional challenges: numerous systems hold financial assets that can be transferred instantly, and the entire infrastructure operates transparently and around the clock.
Therefore, we cannot simply judge the debate between Vitalik and Shapiro by asking whether AI can destroy Bitcoin. The more urgent question is whether attackers will gain a decisive advantage before defensive tools, audit standards, and network infrastructure are strengthened.