AI Disclosure

Exploring designs of AI disclosure, informing consumers whether and how AI-generated content are produced.

Introduction

AI disclosure focuses on informing consumers whether and how AI-generated decisions or content are produced. As AI-generated and co-created content becomes indistinguishable from human-authored work, accurate disclosure is critical for transparency, proper attribution, and helping consumers calibrate their trust in the content.

Problem

Despite emerging policy mandates such as the EU AI Act and South Korea’s AI Basic Act, current disclosure practices remain underdeveloped and poorly aligned with social and institutional norms.

Team

Two user researchers (independent, manager), a UI/UX designer, and a data scientist.

Objective

We examined how AI disclosure should be designed to answer the following questions:

  1. Ownership: how disclosure practices can align with authors’ psychological and legal ownership in co-created work
  2. Accountability: how disclosure shapes understandings of who is responsible for the risks and potential harms associated with AI-generated content
  3. Trust: how different disclosure strategies influence consumer perceptions, trust, and reliance on AI-generated media.

Method

We explored how to develop and design effective AI disclosures across a range of content types (e.g., programming and writing) and modalities (e.g., video, audio, and text).

To do this:

  • We conducted user interviews with key stakeholders, such as product teams developing disclosures, end users who would use them, and attorneys responsible for mandating them.
  • We also held participatory co-design sessions, collaborating directly with these groups to create disclosures that are both practical and helpful.
  • We organized an academic workshop at CHIWORK 2025 to examine how people from diverse backgrounds study and approach AI disclosures. For those interested in learning more, please refer to this Medium blog post.

Deliverables

  1. Patent Filing (Under Review): Method for Tracking AI Authorship of Source Code

  2. AI Attribution Toolkit: We created an online toolkit that creates a disclosure statement that not only shows the presence of AI involvement, but also how AI was used.

  3. Publications:

← Back to Projects