Regulators and policymakers are grappling with the challenges posed by AI-generated content. From the from the rise of deepfakes creating non-consensual imagery of people to bots spreading disinformation and manipulating public discourse, the legal and ethical issues posed by synthetic media are vast. There is thus the growing need to differentiate between AI-generated synthetic content and human content, and several different approaches are being developed and tested to detect how such content is generated as well as its origin. In this blog post, we provide an overview of different approaches to AI-generated content detection, especially watermarking and its effectiveness.

Key Takeaways: