Why labeling authentic media matters

In the push for transparency in the digital age, AI safety best practices are encouraging platforms to label AI-generated content, helping consumers identify when AI has been used. Various governmental and non-governmental organizations are advocating for this, including the EU AI Act, The White House’s Executive Order, and the Partnership on AI. While labeling AI content is an important step toward transparency, it alone is not enough to protect consumers from being misled online.

The limitations of AI labeling

There are two key reasons why relying on the labeling of AI content, while necessary, falls short of ensuring content authenticity:

1. AI without labels: It will always be possible for malicious actors to utilize text-to-image and video models without labels indicating such. This is a result of the time required for ecosystem adoption of Content Credentials, open source models where labels can be removed, and models specifically designed for malicious use. As a counter-balance, consumers will need a reliable way to verify that media is original and untouched from the organizations they care about. By properly labeling authentic content, we can immediately increase transparency and credibility in digital media.
2. Cheapfakes remain a threat: While AI-generated media grabs attention, traditional forms of media manipulation—such as simple edits using basic tools or altering metadata (e.g., date, time, and location)—continue to deceive individuals, companies, and even governments. A recent study highlighted that despite the rise in AI content, conventional media manipulation remains a serious issue.

Content Credentials: enhancing transparency in authentic media

To tackle these challenges, platforms are increasingly offering more transparency around the origins of authentic (non-AI) content. One of the most promising solutions is the use of Content Credentials, an interoperable specification  developed by the Coalition for Content Provenance and Authenticity (C2PA). These credentials provide information about where media comes from, any changes made to it, and whether it’s genuine.

Truepic, as a co-creator of the C2PA, has been a leading advocate for this universal transparency. We believe that allowing creators to showcase the full history of their work is crucial for fostering trust in digital spaces.

Industry progress

Recently, two major social media platforms have made significant strides in labeling authentic content:

• LinkedIn in addition to AI, the platform is now in the process of rolling out display details for authentic images and videos. With Content Credentials, users will begin seeing the provenance and history of the media they encounter.
• YouTube has introduced labeling for authentic streaming videos. This marks an exciting advancement, allowing any authentically captured video aligning to the C2PA specification to be recognized as such by YouTube. This breakthrough has far-reaching implications for industries that depend on transparent content distribution.

These initiatives represent meaningful steps toward a more transparent online experience. Truepic is proud to support some of the first implementations of Content Credentials for non-AI content. Recently, we captured the first authentic video on YouTube using Content Credentials.

^{First authentic video with Content Credentials on YouTube. “How this content was made” highlights that it was captured using Truepic technology, following the C2PA Open Specification.}

Looking Forward

Truepic is focused on enabling capture devices, publishers, and everyone in between with Content Credentials through our Enterprise C2PA solution. We make it extremely easy to implement content authenticity at scale, with nearly a decade of experience powering hundreds of customers on content authenticity solutions. If you’re interested in learning more, please reach out to us here.