Responding to Account Protection in the Mitos Era… DualOS Joins Anthropic’s Security-Focused AI Program

DualAuth (CEO John Jong-hyun Woo) announced on the 15th that it has joined Anthropic’s “Cyber Verification Program (CVP)” as a participating company. In an environment where high-performance AI—such as Anthropic’s top-tier model “Claude Mythos”—is used to discover vulnerabilities and analyze attack paths, the company aims to proactively assess its authentication products and customer services using AI of the same caliber.

Anthropic’s CVP is a program that grants security companies and experts who have passed a separate qualification process access to advanced cybersecurity AI features that are otherwise restricted to general users. Participating companies can utilize Anthropic’s latest AI models for advanced security research, such as malware analysis, vulnerability detection, and attack path verification.

DualAuth plans to use the access to security-specific AI secured through its participation in the CVP to inspect for authentication vulnerabilities in its passwordless authentication product line, “AutoPassword,” as well as to assess vulnerabilities in its clients’ services. The company aims to enhance the security of its products and its clients’ services by proactively identifying flaws in authentication protocols and implementation code, and by verifying attack paths from an attacker’s perspective in real-world service environments.

AutoPassword, developed by DualAuth, is a technology recognized for the superior security of mutual authentication and the convenience of automatic passwords, and has been adopted as the international standard X.1280 (Passwordless Mutual Authentication) by the International Telecommunication Union (ITU-T). Instead of users entering a password, the service presents a new automatic password to the user each time, which the user then verifies and approves on their smartphone. This eliminates the burden of entering or remembering passwords and shifts the responsibility for authentication from the user to the online system. Most importantly, since there is no fixed password for the user to remember or enter, there is no risk of credentials being compromised by phishing sites, and the system is designed to prevent credential stuffing and brute-force attacks that rely on lists of leaked passwords.

In fact, while large-scale AI-powered credential attacks have recently targeted domestic banks one after another, a bank that implemented AutoPassword suffered no damage from these attacks, proving its defensive capabilities even in an AI-based attack environment.

John Jong-hyun Woo, CEO of DualAuth , stated, “In an era where AI on par with Mitos is falling into the hands of attackers, protecting accounts and authentication is the starting point for cybersecurity.” He added, “Using the security-specialized AI secured through CVP, we will proactively inspect AutoPassword authentication vulnerabilities and customer services to shorten verification cycles and increase the depth of our inspections.”