DIGITAL WATERMARKING AND THE FIGHT AGAINST DEEPFAKE CONTENT

The rapid advancement of deepfake technology has created serious challenges to the authenticity and security of digital media. Deepfake content generated using artificial intelligence has increasingly been misused for misinformation, fraud, and identity manipulation, reducing public trust in digital platforms. This study examined digital watermarking as a proactive defense mechanism against deepfake manipulation by embedding imperceptible yet robust identifiers into multimedia content. A simulation-based approach was applied to evaluate watermark performance under several deepfake generation techniques, including FaceSwap, DeepFakes, and NeuralTextures. The evaluation focused on robustness, imperceptibility, and resistance to common post-processing operations such as compression, noise addition, and format conversion. Experimental results demonstrated that the proposed watermarking method maintained reliable detection performance while preserving visual quality, as indicated by acceptable Peak Signal-to-Noise Ratio (PSNR) values. Furthermore, the embedded watermark showed resilience against adversarial attempts to remove or distort the hidden information. These findings suggest that digital watermarking can serve as an effective complementary approach to deepfake detection systems. Overall, this research highlighted the importance of digital watermarking in strengthening media integrity amid the growing sophistication of synthetic media technologies