The integration of artificial intelligence into surveillance technologies is transforming exceptional investigative tools into persistent systems with profound implications for fundamental rights and governance. This article examines how AI-enabled spyware, lawful trojans, and automated monitoring practices challenge core principles of proportionality, necessity, and oversight. By combining insights from cybersecurity, digital forensics, and AI governance, the analysis shows how automation expands surveillance scope, fragments accountability, and renders traditional oversight mechanisms increasingly ineffective. AI-assisted targeting and inference undermine the ability to assess proportionality ex ante, shifting governance from preventive control to reactive damage management. The article argues that without enforceable limits, explainability requirements, and forensic auditability, AI-enabled surveillance risks evolving into a structural governance failure, eroding public trust and institutional legitimacy. It concludes that effective AI governance must embed rights protection, oversight, and accountability into surveillance architectures to prevent security technologies from undermining the democratic foundations they claim to protect.