Deepfake voice technologies have heightened concerns about security, privacy, and trust in digital communication. This paper provides a detailed review of how deepfake voices are created and the methods used to detect and prevent them. Voice recognition systems that try to catch deepfake audio are examined, including how well they work, their weaknesses, and what needs to be improved. The ethical issues of using deepfake voices, along with the risks they create for cybersecurity, fraud prevention, and media trust, are also discussed. It concludes that the outlook is toward a chain of advancements in recognizing deepfake voice and further resilient detection frameworks for protection against misuse of the powerful technology.