Crypto
Home›Crypto›Market Structure›Crypto firms seek frontier AI access, but security too…
Crypto firms seek frontier AI access, but security tools remain uneven
Coinbase said in June it gained access to Anthropic’s restricted Mythos model, while other major firms, including Binance, have been trying without getting the most frontier model.
Crypto executives say they are still trying to secure access to the most powerful, frontier AI models to help defend their code from attacks, but access is currently uneven across the industry. Cointelegraph reports that only a select few crypto players have been able to obtain such tools.
Coinbase, for example, said in June that it secured access to Anthropic’s restricted Mythos model. Cointelegraph also notes that Zcash’s Zooko Wilcox said Anthropic used the Mythos model to audit the Zcash protocol at the request of Shielded Labs.
Binance’s chief security officer Jimmy Su said the exchange has been attempting to make progress, including talks with other crypto exchanges and its investors, but has not obtained the most frontier AI model like Mythos. Cointelegraph adds that the uneven availability is creating what executives describe as a security divide, with potential consequences for firms facing increasingly sophisticated, AI-assisted threats.
Cointelegraph reports that while model developers such as Anthropic and OpenAI restrict their most cyber-capable models from the public, growing capabilities in open-source alternatives could leave crypto firms without the strongest defensive tools. Anthropic says Mythos 5 uses the same underlying model as its publicly available Fable 5, but without safeguards that restrict sensitive cybersecurity work, and OpenAI uses a tiered approach for GPT-5.5 access.