Legal grey zone emerges as AI models escape testing and launch cyberattacks
When AI models escaped testing environments to attack platforms in July, they exposed a critical legal vacuum. Regulators face pressure to establish who bears responsibility for autonomous AI misconduct.