AI has been a ‘net negative’ for crypto: Phemex CEO
Phemex CEO Federico Variola told Cointelegraph’s Chain Reaction that artificial intelligence has been a net negative for the crypto industry, arguing it has siphoned liquidity away from crypto, empowered attackers, and raised cybersecurity costs that could push projects toward greater centralization. His remarks follow Phemex’s own AI-focused transformation announced in February and come amid a spate of AI-linked security incidents in the space.
Why It Matters
Variola’s assessment highlights tensions between AI-driven innovation and emerging security risks in crypto: if AI both attracts capital away from crypto and makes attacks cheaper or more effective, smaller teams may be forced to centralize or scale back, reshaping the industry’s decentralization goals. The debate also reflects wider industry concern as high-profile exploits and expert warnings underscore both offensive and defensive uses of AI.
Key Facts
- person: Federico Variola, CEO of Phemex
- platform: Cointelegraph's Chain Reaction (interview)
- claim: AI has been a net negative for crypto by diverting capital, empowering attackers, and increasing cybersecurity costs
- company-action: Phemex announced an AI-focused transformation in February to embed AI across product development and operations
- exploit-example: $116 million in Bitcoin drained from over 5,200 addresses tied to a Coldcard hardware wallet flaw (July), widely believed to involve malicious use of AI
Phemex CEO Federico Variola told Cointelegraph’s Chain Reaction that he views artificial intelligence as having an overall negative effect on the crypto sector. Variola said AI has siphoned liquidity away from crypto toward other AI-focused ventures, while simultaneously giving bad actors new tools to probe and exploit protocols. He cautioned that rising defensive costs are squeezing smaller development teams and could nudge projects toward more centralized models.
Variola’s comments come after his own exchange outlined an AI strategy in February that planned to integrate the technology across product development and internal operations. Despite those in-house plans, his recent remarks focused on industry-wide impacts rather than Phemex’s commercial use of AI. He argued that many of the fixes being developed to counter AI-enabled threats tend to favor centralized solutions over the decentralized ideals that underpin much of crypto.
Security incidents cited by industry observers bolster some of his concerns. In July, attackers removed about $116 million in Bitcoin from more than 5,200 addresses linked to a Coldcard hardware-wallet vulnerability that has been widely reported as likely discovered through malicious AI activity. Coinkite’s CEO warned developers that AI-assisted code review can now surface vulnerabilities faster than human reviewers, and Variola said that small protocol teams will soon need much larger cybersecurity budgets to keep pace.
Not all experts see only downside: some security researchers say AI can also strengthen defenses even as it makes attacks more sophisticated. Variola acknowledged practical uses for AI-driven agents, for instance in helping investors build portfolios or refine trading decisions, but he does not expect such agents to fully replace human judgment on final trades. The exchange between defensive and offensive AI capabilities suggests the industry will continue grappling with how to deploy the technology without compromising decentralization or leaving smaller players exposed.
Keep Reading
XRP sinks 10% as the Clarity Act fails and bitcoin slides toward $76,000
ETH wallet exploit backfires as MEV bot captures $7.7M, Kelp freezes address
Standard Chartered says Arbitrum could outperform Bitcoin, Ether through 2030
