I-Corps: Translation Potential of a Machine Learning-based Watermark for Digital Intellectual Property Protection

The broader impact/commercial potential of this I-Corps project is the development of watermarking technology designed to protect audio content from unauthorized use and deepfake manipulation. Currently, the use of unauthorized audio and deepfake content, which is audio that has been digitally altered, is a growing problem. The proposed technology may protect audio creations and discourage unethical practices by equipping artists, producers, and copyright holders with the means to safeguard their work. Ultimately, the proposed technology advances scientific and technological understanding in digital rights management, significantly impacting society and the commercial sector by ensuring audio content's authenticity in a digital age. In addition, it may be used as a tool for copyright enforcement, offering an efficient way to protect intellectual property and ensure fair compensation for creators as well as to verify authenticity. This I-Corps project utilizes experiential learning coupled with first-hand investigation of the industry ecosystem to assess the translation potential of the proposed technology. It is based on the previous development of audio watermarking technology using adversarial and machine learning techniques. The technology is focused on injecting audio watermarks that are imperceptible to human listeners yet can be easily identified by proprietary detection algorithms with the goal of ensuring the integrity and authenticity of digital media. Additionally, this technology represents a pivotal step towards establishing more resilient frameworks against the misuse of artificial intelligence (AI) in content creation. It may provide a foundation for future research in secure content management and strengthen defenses against AI misuse in content creation, providing essential tools for copyright protection. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.