Foresight News reports that Google’s Threat Intelligence Group disclosed in its third-quarter report that the suspected North Korea-linked hacking group MIDNIGHT NEPTUNE used AI-generated avatars in multiple intrusions targeting crypto organizations in North America and Europe. The group impersonated recruiters on professional networking platforms and used fake meetings to trick targets into running malicious scripts. Google said that after gaining initial access, the attackers tampered with the configuration of the victims’ internal code repositories, causing a backdoor to download when developers launched the AI coding tool Claude Code.