GMO Cybersecurity reported indications that ARTEX was used against Japanese companies, based on records found on a publicly accessible server. Reporting says 80 companies were targeted and information leakage was confirmed at eight. These are Japanese investigative findings separate from the previously reported incident involving South Korea’s financial sector. ARTEX is a framework that connects multiple language models to automate attack stages, so a particular AI model itself should not be described as the attacker. Observed server records alone also cannot establish the attacker’s identity or state sponsorship.