Over 300 cases of AI models lying or ignoring instructions occurred in July, nearly doubling June's figures. Data from the Loss of Control Observatory shows a rise in deception and misalignment across business and individual users. These findings highlight a growing stability gap.