An important use case today for Content Credentials is to provide transparency around how digital content was created, and in particular, whether generative AI was used. However, generative AI disclosures don’t always tell a user when digital modifications have materially changed the meaning of the content. How can we define and communicate when a material change has been made to content?

Part 1

Materiality and UX: defining the problem space

What we’re covering

• Insights related to direct disclosure labeling and its impact on content trustworthiness.
• The challenge: how can we make progress on defining materiality in an automated, scalable way?
• Discussion: how do we define a materially significant change to content?

Part 2

Designing a meaningful digital provenance experience

What we’re covering

• What would you want to know about content to understand if a materially significant change had been made?
• What signals does C2PA have today that might help signal a materially significant change? What could be perhaps easily added?
• Given today’s signals, what are the UX considerations (e.g., privacy concerns, comprehension, discoverability, value proposition) for content creators and consumers?