Crypto Exchanges Prepare for AI Capable of Detecting Vulnerabilities
Major cryptocurrency exchanges and custodial platforms, including Coinbase (NASDAQ:COIN) and Binance, are preparing their systems for the emergence of powerful artificial intelligence models capable of identifying and exploiting software vulnerabilities, according to The Information.
Anthropic’s Mythos model, described by the AI developer as potentially dangerous due to its ability to uncover weaknesses in security systems, has prompted crypto platforms to accelerate defensive upgrades.
Coinbase Chief Information Security Officer Philip Martin stated that the exchange remains in close communication with Anthropic regarding Mythos, which he described as a tightly restricted model unavailable to the general public.
Martin explained that Mythos and similar future AI systems could enable deeper software and infrastructure testing at scale, accelerating both the evolution of cyber threats and digital defense mechanisms.
Anthropic granted early access to Mythos for testing to a select group of organizations, including major banks. However, large cryptocurrency firms were not included in this initial testing group.
See also: "Coinbase: Stablecoin compromise paves the way for U.S. crypto legislation"
Українська
Русский
English

