By Alessandra Sala, Chair of the AI and Multimedia Authenticity Standards Collaboration
Globally, people’s shared understanding of evidence, creativity, and trust is being transformed by artificial intelligence (AI). Today, AI-generated and AI-edited media are no longer experimental edge cases.
By now, synthetic media is a foundational layer of information exchange that is reshaping traditional benchmarks of what is “real.”
For the AI-native generation, synthetic media is part of the digital landscape. It informs how they create, share, learn, and participate in public life.
While the deep embrace of AI opens exciting frontiers for access and expression, it also raises urgent questions: What does authenticity mean when content can be generated instantly? How can people know what to trust? And how can creative freedom be protected alongside digital safety?
To help answer these questions, the leading industry partnership on AI and multimedia authenticity standards – established in 2024 by the World Standards Cooperation (IEC, ISO and ITU) – is fostering active collaboration among early-career AI researchers worldwide.
The Young Researcher Associate Programme under the AI and Multimedia Authenticity Standards Collaboration, brings together promising research talent from universities to tackle the fundamental challenges of trust, provenance, and rights protection in the era of ubiquitous AI content. The one-year Young Researcher Associate Programme reflects a simple idea: the future of content integrity cannot be assured without design input of the generation of researchers who living most deeply inside today’s digital ecosystems.
Next week at the AI for Good Global Summit, these researchers will present their work to build trust in digital content. Discover the event here.
Three projects for a trustworthy information ecosystem
This year’s cohort centers on three complementary, student-led projects.
Each addresses a distinct layer of the authenticity challenge: the human cognitive layer, the technical explanation layer, and the global policy layer. Together, they bridge the gap between theory and practice to offer a holistic vision for multimedia authenticity in action.
1. STOP&SCAN: Engaging critical thinking
A project led by the programme’s social science team addresses the human cognitive dimension of digital media integrity. As generative AI fuels deepfakes and voice scams, STOP&SCAN applies psychological inoculation and trust calibration research to strengthen critical judgment.
The five-step framework encourages users first to pause (or Stop) and then assess the Source, Content, and Alignment of information, and finally Reflect before acting.
The team has reinforced this approach with three supporting tools: Amito, a friendly robot mascot for the AI for Good Digital Media Integrity Toolkit; an interactive learning website; and a comic series that reinforces key messages through engaging stories.
2. AMITO: An agentic toolkit against deepfakes
Existing detection tools are often too technical or difficult for citizens to use. A project led by the technical team delivers accessible media analysis through everyday messaging platforms, allowing users to readily forward suspicious images or videos for analysis.
The AMAS Multimedia Integrity Toolkit, or AMITO, is already available on Telegram and WhatsApp.
The AI orchestrator system examines provenance signals, metadata, manipulation indicators, and open-source intelligence evidence, with model-agnostic architecture selecting tools anonymously, evaluating evidence, and educating users interactively. AMITO gives its findings in plain language. Rather than simply deeming content “fake” or “real”, AMITO explains the evidence, uncertainties, verdict, and what the user can do next.
3. Policy-as-Code: Mapping content across borders
AI content moves globally at the speed of a click, but laws stop at national borders. This jurisdictional mismatch creates massive compliance challenges for creators and platforms alike. Led by the policy and business team, the Policy-as-Code project bridges this gap by mapping jurisdictions geographically and with their current regulatory status. At the center of the project is an interactive regulatory map. Jurisdictions are plotted geographically and color-coded by regulatory status: enacted and enforced, draft, voluntary standard, or regulatory gap. Each jurisdiction is clickable, allowing users to explore the rules that would apply to a piece of AI-generated content the moment it enters that legal environment. For the first time, the journey of AI-generated content across borders can be traced against the patchwork of laws it passes through.
The map is also the foundation for a more ambitious goal: a machine-readable compliance passport built on C2PA. This passport is designed to travel with content across platforms, borders, and regulatory regimes that were not originally designed to communicate with one another. This project helps policymakers, platforms, creators, and businesses see where legal obligations align, where they conflict, and where gaps remain. It translates regulatory complexity into practical infrastructure, making compliance more transparent, interoperable, and scalable. In a global information ecosystem, digital trust must be able to travel. Policy-as-Code shows how that future might begin.
Collaboration for an authentic future
The Young Researcher Associate Programme is a model for how digital trust should be built: collaboratively, internationally, and across generations.
By converting abstract principles into practical tools, these young researchers are showing the world how verification can become accessible and how trust can be designed into our digital fabric without shutting down innovation.
Let’s connect at AI for Good
Want to meet the young researchers behind these projects and explore the future of multimedia authenticity?
Join us on 9 July at the AI for Good Global Summit in Geneva, Switzerland, where the AI and Multimedia Authenticity Standards Collaboration will spotlight new research, practical tools, and emerging approaches to building trust in digital content:
Images credit: AI and Multimedia Authenticity Standards Collaboration
