coindesk.com
Network News
MYTHOS CHALLENGES CRYPTO SECURITY: Mythos, the new AI model from Anthropic that has sparked fear and confusion in traditional tech and finance, is also driving a massive shift in how the crypto industry thinks about security. For years, decentralized finance has focused its defenses on smart contracts. Code is audited, vulnerabilities are cataloged, and many common exploits are well understood. But Mythos, a model designed to identify and chain together weaknesses across systems, is pushing attention beyond code and into the infrastructure that supports it. “The bigger risks sit in infrastructure,” said Paul Vijender, head of security at Gauntlet, a risk management firm. “When I think about AI-driven threats, I’m less concerned about smart contract exploits and more focused on AI-assisted attacks against the human and infrastructure layers.” That includes key management systems, signing services, bridges, oracle networks and the cryptographic layers that connect them. These components are less visible than smart contracts and are often outside the scope of traditional audit. In fact, this month, web infrastructure provider Vercel, used by many crypto companies, disclosed a security breach that may have exposed customer API keys, prompting crypto projects to rotate credentials and review their code. Vercel traced the intrusion to a compromised Google Workspace connection via the third-party AI tool Context.ai, which an employee used. Mythos belongs to a new class of AI systems built to simulate adversaries. Instead of scanning for known bugs, it explores how protocols interact, testing how small weaknesses can be combined into real-world exploits. That approach has drawn attention beyond crypto. Banks like JP Morgan are increasingly treating AI-driven cyber risk as systemic and are exploring tools like Mythos for stress testing. Earlier this month, Coinbase and Binance both reportedly approached Anthropic to test Mythos. Early findings from models like Mythos have identified weaknesses in the behind-the-scenes systems that keep crypto platforms secure, including the technology that protects keys and handles communication between systems. — Margaux Nijkerk Read more.
AAVE’S $300M RECOVERY EFFORT: In the often-fractured world of decentralized finance, crises tend to expose fault lines. This time, they’re also revealing an unusual level of coordination. Aave, one of DeFi’s largest lending protocols, is at the center of a broad recovery effort following losses tied to the Kelp DAO exploit, drawing in capital and credit commitments from across the industry. The effort, informally dubbed “DeFi United,” had raised about $301 million in commitments as of Monday, according to its website, with much of the capital still pending governance approval. The exploit, which rippled into rsETH markets and created risk across lending positions on Aave, has prompted what is shaping up to be one of the most coordinated industry responses to a DeFi incident. “There’s a shared priority around supporting users and restoring normal market conditions,” an Aave Labs spokesperson told CoinDesk. “Many of these participants are deeply connected to DeFi, whether through infrastructure, capital, or user access, and have a direct interest in ensuring markets function as expected.” At the core of the effort is Aave itself. A governance proposal outlines a plan for the DAO to allocate up to 250,000 $ETH as part of the recovery. Founder Stani Kulechov separately indicated he would donate 5,000 $ETH personally. Other contributors within Aave’s orbit are also stepping in, including Aave’s Emilio Frangella (500 $ETH), BGD Labs’ Ernesto Boado (100 $ETH), BGD Labs (250 $ETH) and KPK’s Marcelo Ruiz de Orlano (100 $ETH). The response has quickly extended beyond Aave, and in some cases began with direct outreach. Following the April 18 bridge hack, Kulechov reached out to Consensys and other ecosystem participants early to help coordinate a response, according to a Consensys spokesperson. The firm, alongside its founder Joseph Lubin, agreed to commit up to 30,000 $ETH in financial support to help advance the recovery and protect users. Sharplink played a strategic advisory role in those discussions, the spokesperson said. — Margaux Nijkerk Read more.
CRYPTO IS FOR AI AGENTS, SAYS ALCHEMY CEO: The modern financial system was never designed for machines. It was built around the constraints of human life: geography, sleep cycles, paperwork and physical presence. But as AI agents begin to act as economic participants, that human-centric design is starting to look less like a feature and more like a bottleneck, said the co-founder of crypto firm Alchemy. “You can argue that crypto was built for AI agents, not humans,” said Nikil Viswanathan, who is also CEO. The mismatch is everywhere. Banks have operating hours because humans do. Payments are tied to countries because people live in them. Credit cards assume physical identity and presence, he said. AI agents operate differently. They don’t sleep. They don’t live anywhere. They don’t walk into banks or carry cards. And increasingly, they don’t just assist with tasks, they transact. “All transactions for agents are online. They’re inherently global,” Viswanathan, who will be speaking at Consensus Miami next month, told CoinDesk in an interview. That’s where crypto starts to look less like an alternative financial system and more like the native infrastructure for a new kind of economic actor, he said. Traditional finance assumes friction. Paying someone in another country involves currency exchanges, intermediaries, delays, fees. For humans, that’s normal; for AI agents, it’s unusable. Agents need to transact seamlessly across borders, at any time, often in tiny increments. They need programmability, direct control over money via code and systems that don’t depend on physical infrastructure or identity. Crypto offers exactly that: a global, always-on financial layer where value moves as easily as data, he said. “Crypto is the global infrastructure for money that agents need,” Viswanathan said. — Will Canny Read more.
