How Anthropic's Mythos Model Is Revolutionizing Crypto Security

The introduction of Mythos, Anthropic's innovative AI model, has sent shockwaves through the traditional tech and finance sectors, and is now prompting a significant paradigm shift in the crypto industry's approach to security. Historically, decentralized finance has focused on bolstering its defenses by scrutinizing smart contracts, cataloging vulnerabilities, and understanding common exploits. However, Mythos, with its ability to identify and exploit weaknesses across entire systems, is redirecting attention towards the underlying infrastructure that supports these contracts. Paul Vijender, head of security at Gauntlet, a risk management firm, emphasized, 'The more significant risks lie in the infrastructure.' He added, 'When considering AI-driven threats, my primary concern is not smart contract exploits, but rather AI-assisted attacks targeting human and infrastructure layers.' These components encompass key management systems, signing services, bridges, oracle networks, and the cryptographic layers that interconnect them. Notably, these elements are less visible than smart contracts and often fall outside the scope of traditional audits. Recently, web infrastructure provider Vercel, which is widely used by crypto companies, disclosed a security breach that may have compromised customer API keys, prompting crypto projects to reevaluate their credentials and code. The breach was attributed to a compromised Google Workspace connection via the third-party AI tool Context.ai, used by an employee. Mythos represents a new generation of AI systems designed to simulate adversarial attacks. Rather than searching for known vulnerabilities, it explores how protocols interact, testing how minor weaknesses can be combined to create real-world exploits. This approach has garnered attention beyond the crypto sphere, with banks like JP Morgan treating AI-driven cyber risk as systemic and exploring tools like Mythos for stress testing. Early findings from models like Mythos have identified vulnerabilities in the behind-the-scenes systems that secure crypto platforms, including the technology that safeguards keys and facilitates communication between systems. According to Vijender, 'I believe AI models are particularly valuable in two areas: first, identifying multi-step exploit chains that historically only come to light after funds have been lost, and second, uncovering infrastructure-layer vulnerabilities that traditional audits often overlook.' This shift is significant in a system built on composability, where DeFi protocols can interconnect and build upon each other's services. DeFi protocols are designed to be interconnected, sharing liquidity, relying on common oracles, and interacting through layers of integrations that are challenging to fully map. This interconnectedness has driven growth but also creates pathways for risk to spread, as seen in recent bridge exploits. 'Composability is what makes DeFi capital-efficient and innovative,' Vijender noted. 'However, it also means that a minor vulnerability in one protocol can become a critical exploit vector with contagion potential across the ecosystem.' Without AI, these dependencies are difficult to trace. With AI, they can be mapped and exploited on a large scale, resulting in a shift from isolated exploits to systemic failures that cascade across protocols. The evolution of AI attacks is still unfolding, with some industry leaders viewing Mythos as an acceleration rather than a turning point. Stani Kulechov, founder of Aave Labs, remarked, 'AI reflects the dynamics already at play in DeFi's adversarial environment.' He added, 'AI models represent an evolution in the tools used to achieve exploits.' From this perspective, DeFi is already built for machine-speed attacks, with smart contracts executing automatically and defenses like liquidation mechanisms and risk parameters operating without human intervention. 'DeFi operates at compute speed, so AI doesn't introduce a new dynamic,' Kulechov said. 'It intensifies an environment that has always required constant vigilance.' Even so, Aave is discovering AI surfacing new categories of vulnerabilities, including issues that human auditors may have previously deprioritized. 'The Mythos paper shows that AI can uncover old bugs that were previously deprioritized,' he said. This breadth still matters in a system where even smaller vulnerabilities can undermine trust or be combined into larger exploits. If attackers can move faster, the question becomes whether defenses can keep pace. For both Gauntlet and Aave, the answer lies in changing the security model itself. Audits before deployment and monitoring after were designed for human-paced threats; AI compresses that timeline. 'To defend against offensive AI, we will need to take an AI-centric approach where speed and continuous adaptation are essential,' Vijender of Gauntlet said. This includes continuous auditing, real-time simulation, and systems built with the assumption that breaches will occur. Aave has already integrated AI into its workflows, using it for simulations and code review alongside human auditors. 'We take an AI-first approach where it adds clear value,' Kulechov of Aave Labs said. 'But it complements, rather than replaces, human-led auditing.' In this sense, AI equips both attackers and defenders. For builders, the long-term effect may be less disruption than divergence. Hayden Adams, founder and CEO of Uniswap Labs, said, 'We haven't tested Mythos yet, but we're genuinely interested in what it and tools like it can do for protocol security.' He added, 'AI gives builders better ways to stress test and harden systems.' Over time, Adams expects the gap between secure and insecure protocols to widen. 'Projects that prioritize security will have greater ability to test and harden systems before launching,' he said. 'Projects that don't will be most at risk.' That may be the real shift. Security is no longer about eliminating vulnerabilities; it is about continuously adapting to a system in which those vulnerabilities are constantly rediscovered and recombined.