Deepfakes have come a long way in less than 3 years. Deepfakes are no longer confined to laboratory experiments and are now evolving into a systemic threat. As of early 2026, the use of AI-generated content has increased by over 900% every year, and seeing is believing has now become a statement that is outdated. This paper discusses how deepfakes are one of the key causes of the crumbling trust in online information systems and digital media. It further explains their role in misleading the public, manipulating politics and causing social deception. The paper then evaluates the current detection mechanisms for deepfakes and discusses some approaches for a better and reliable trust in digital media.