In the digital age, AI-driven surveillance technologies, including predictive policing and facial recognition systems, have shifted from speculative fictional concepts to pressing societal concerns. These systems have provoked extensive debate among policymakers, legislators, and human rights advocates due to their capacity to infringe upon privacy and civil liberties. This paper is part of a broader research initiative that examines the discourse surrounding AI-powered surveillance; while our broader research focus includes case studies of predictive policing, facial recognition, and social media monitoring, this specific paper will examine predictive policing. Employing Critical Discourse Analysis (CDA), the study interrogates the competing narratives of technological advancement versus privacy protection, alongside the tension between industry self-regulation and governmental oversight. Central to this inquiry are the ethical implications of AI surveillance, including its capacity to enhance authoritarian control, exacerbate systemic biases, and facilitate abuses of power, particularly within law enforcement, population management, and the regulation of human movement. Drawing on case studies and stakeholder perspectives, this research elucidates the ethical and social dilemmas inherent in AI surveillance, emphasizing the necessity for heightened scrutiny, regulatory intervention, and a critical examination of the discourse that underpins surveillance practices. The paper further proposes a framework of surveillance heuristics to deconstruct the justificatory language embedded in digital texts—such as terms of service—that normalize and legitimize surveillance. Ultimately, the study advocates for the responsible integration of AI technologies, emphasizing the importance of safeguarding privacy while acknowledging the potential benefits of innovation in the digital age.