The swift progression of Artificial Intelligence has changed the character of evidence in judicial proceedings. AI-generated content, including deepfake videos, synthetic audio, and AI-made digital content, poses serious and new challenges to the basic rules of admissibility, authenticity, and reliability of evidence. Although Indian law has been updated through the Bharatiya Sakshya Adhiniyam, 2023, which formally acknowledges electronic records as evidence, it does not specifically deal with the unique problems of AI-generated content. This paper looks at how Indian courts can deal with this problem before it gets worse. This paper examines the legal position of AI-generated evidence before Indian courts by looking at existing laws, court decisions and legal rules on electronic evidence. It also studies the challenges created by such evidence, including difficulties in authentication, the widely discussed “black box” problem of AI systems, break in chain of custody, and the potential for deliberate misuse. A comparative study of approaches adopted in jurisdictions such as the United States, the European Union, and the United Kingdom is conducted to identify international best practices. The paper argues that the existing Indian legal framework is not equipped to handle the specific dangers of AI-generated evidence and calls for clear changes, including explicit statutory recognition, better verification methods, required disclosure rules, and training for judges and courts. The study concludes that while AI can genuinely help justice, its uncontrolled use can destroy the very idea of truth that courts depend on.