Building Safe and Fair Generative AI with Content Provenance | John Collomosse, Adobe Research
Follow SAI Conferences on Linkedin: / saiconference
Conference Website: https://saiconference.com/FTC
Prof. John Collomosse is a principal scientist at Adobe Research, where he leads research for the Content Authenticity Initiative (CAI) and two cross-industry task forces within the C2PA open standards body for media authenticity. He is a professor at the University of Surrey, where he is the founder and director of DECaDE, the UKRI Research Centre for the Decentralized Creative Economy. His research focuses on media provenance to fight misinformation and online harms, and on improving data integrity and attribution for responsible AI.
In this keynote presentation, John Collomosse delves into the world of generative AI and the critical role of content provenance in ensuring transparency, fairness, and accountability. He explores how generative AI models, which have been around for over a decade, have rapidly evolved to create highly diverse and realistic content—ranging from text-to-image to video generation. Collomosse highlights the driving forces behind these advancements, such as the massive amounts of data used to train AI models and the importance of understanding data sources and consent in training processes.
Throughout the talk, Collomosse addresses the challenges posed by AI, including biases in training data, memorization of copyrighted content, and the growing concern of misinformation and fake news. He also introduces innovative solutions such as the Coalition for Content Provenance and Authenticity (C2PA), which is working on creating an open standard for content credentials—helping users verify the origins of digital content. He further discusses the importance of metadata, watermarking, and fingerprinting to ensure that content’s provenance is preserved across platforms, enabling users to make informed decisions about what they see online.
Collomosse concludes by sharing insights into future developments in AI consent management and the potential for provenance-based models to unlock new value creation opportunities for content creators.
Key Takeaways:
Understanding generative AI and its ability to create realistic content.
The impact of training data on AI model behavior and performance.
The importance of content provenance to fight misinformation.
How standards like C2PA and technologies like watermarking can help ensure transparency in AI-generated media.
Exploring the future of AI consent management and new value creation models.
Stay tuned for a comprehensive discussion on how we can ensure the safe and fair use of AI while protecting the rights of creators!