Major crypto exchanges, including Coinbase and Binance, are accelerating their efforts to prepare for advanced AI models, notably Anthropic’s Mythos, which specializes in identifying remote code execution vulnerabilities. This initiative comes in the wake of increasing concerns over AI-driven exploits, prompting these firms to seek access to Mythos for testing their platforms. Alongside this, Anthropic’s Project Glasswing aims to partner with organizations to enhance proactive software security in an increasingly AI-centric landscape.

Mythos: Claude Mythos Preview is Anthropic’s most advanced frontier AI model to date, demonstrating exceptional abilities in autonomous security research and identifying vulnerabilities in software, browsers, and operating systems. Released in limited preview to trusted partners to mitigate risks, it powers defensive efforts like Project Glasswing. Crypto exchanges are urgently preparing infrastructure to integrate or test Mythos for enhanced defenses against AI exploits.
Binance: Binance is the world’s leading blockchain ecosystem and digital asset exchange, providing trading, wallets, and AI-integrated tools like Binance AI Pro for autonomous trading agents and prediction markets. It has incorporated multiple AI models for trading strategies and skills across payments and earn products. Binance is among major exchanges preparing for Anthropic’s Mythos model, building on prior use of Anthropic models for vulnerability testing.
Coinbase: Coinbase is a leading U.S.-based cryptocurrency exchange platform offering trading, custody services for ETFs, layer-2 solutions like Base, and innovations such as crypto-backed mortgages and fast stablecoin settlements. It recently received conditional approval for a national trust charter and is expanding into prediction markets. Coinbase is negotiating with Anthropic to access the Mythos AI model to bolster platform security against advanced threats.
Anthropic: Anthropic is an AI research company developing safe frontier large language models in the Claude family, emphasizing alignment and capabilities in coding and reasoning. It recently launched Claude Mythos Preview, a highly capable model with strong agentic skills, and Project Glasswing to secure critical software. Anthropic’s Mythos is prompting crypto exchanges like Coinbase and Binance to race for access due to its cybersecurity prowess.

Model Capabilities: Mythos excels at finding remote code execution vulnerabilities overnight without formal security training.
Industry Initiative: Anthropic’s Project Glasswing partners with organizations to leverage Mythos for proactive software security in the AI era.
Security Preparations: Crypto firms including Coinbase and Binance seek Mythos access to test platforms amid concerns over AI-driven exploits.