Frenos emphasized that its cyberattack tests used supplied information and simulated scenarios without proving that artificial intelligence models could independently identify targets or execute cyberattacks.