AI Agents Will Increasingly Exploit Weak Authentication by Automating Credential Theft and Compromising Authentication Communication Channels By 2027, AI agents will reduce the time it takes to exploit account exposures by 50% according to Gartner, Inc. "Account takeover (ATO) remains a persistent attack vector because weak authentication credentials, such as passwords, are gathered by a variety of means including data breaches, phishing, social engineering and malware," said Jeremy D'Hoinne, VP