Fastly's Enhanced Commitment to Content Provenance 

The fight against misinformation has never been more critical. As deepfakes become more sophisticated and disinformation campaigns more prevalent, especially due to the rise of AI, the need for robust content authentication solutions grows exponentially. At Fastly, we've been working alongside the Coalition for Content Provenance and Authenticity (C2PA) since its early days, and today we're excited to share how we're taking our commitment to content provenance to the next level with support in Fastly Image Optimizer.

Introducing C2PA Support in Fastly Image Optimizer

We've now enhanced Fastly Image Optimizer with native C2PA support through a new metadata parameter option: `metadata=c2pa`. This parameter value:

• Preserves existing C2PA manifests from source images while adding new provenance data about Fastly's processing

• Signs C2PA manifests cryptographically to maintain the chain of trust

• Attaches complete provenance information to optimized output images

• Maintains compatibility with existing copyright preservation options through combined parameters like `metadata=c2pa,copyright`

This means that as images flow through Fastly's global network and undergo optimization, compression, or format conversion, their authenticity trail remains intact and verifiable. Every transformation becomes part of the documented history rather than a break in the chain.

Why This Matters: Real-World Impact

The implications of this integration extend far beyond technical specifications. Consider these scenarios:

• News Organizations.  A photojournalist captures an image with a C2PA-enabled camera. As that image moves through editorial systems, gets resized for different platforms, and is delivered via Fastly's CDN, the complete provenance chain remains intact. Readers can verify not just who took the photo, but exactly how it was processed and delivered.

• Content Creators.  An artist creates digital artwork with embedded C2PA signatures. When their work is shared across social platforms, optimized for mobile viewing, or integrated into larger compositions, the creator's attribution and the work's authenticity remain cryptographically verifiable.

• E-commerce Platforms.  Product images maintain their provenance as they're automatically optimized for different device types and network conditions, helping consumers verify the authenticity of what they're seeing.

Fastly Image Optimizer Meets C2PA: Preserving Trust Through Transformation

We previously wrote about the promise of C2PA technology and how new technical standards could create a more trustworthy internet. At the time, C2PA was an ambitious vision—a way to cryptographically bind the history of digital media so users could make informed decisions about the content they consume and share. Today, that vision is becoming reality. As the internet has evolved, so has our understanding of the misinformation challenge. We've seen how manipulated media can spread faster than fact-checkers can respond, and how the lack of provenance information makes it nearly impossible for users to assess the authenticity of the content they encounter daily.

One of the most significant challenges in implementing content provenance has been maintaining authenticity metadata as media moves through complex delivery pipelines. When images are optimized, resized, or transformed for different devices and contexts, critical provenance information is often lost, breaking the chain of trust that C2PA aims to establish.

We're solving this problem directly within Fastly Image Optimizer, our service that optimizes images as they pass through the Fastly network. Historically, our optimization process stripped away metadata by default, prioritizing performance and file size reduction. Metadata preservation has proven to be a valuable addition to Fastly Image Optimizer, striving to balance the web performance benefits of only preserving what is absolutely necessary (to minimize file size) for some images. C2PA is the latest enhancement in our support for using metadata.

The Broader Ecosystem and What’s Next

Our C2PA integration represents more than just a feature addition—it's part of a broader commitment to building infrastructure that supports a more trustworthy internet. By implementing C2PA at the edge, we're helping to normalize content provenance as a standard part of media delivery. While we're proud of this technical achievement, we recognize that widespread adoption of content provenance technology requires more than just robust infrastructure. It demands continuous efforts and focus through industry collaboration, user education, and more, with a commitment to privacy protection and accessibility for creators and publishers of all shapes and sizes.

This work aligns with the expanding C2PA ecosystem. The standard is gaining traction beyond static images, with initiatives like C2PA for on-demand and live video bringing real-time authentication to streaming content, with applications across news, entertainment, and social media platforms.

Join Us in Building a More Trustworthy Internet

The challenge of misinformation requires sustained effort from across the technology ecosystem. By integrating C2PA support into Fastly Image Optimizer, we're providing the infrastructure that makes content provenance practical at internet scale. Technology alone isn't enough, though. We need continued collaboration between platforms, creators, publishers, and users to build an internet where authenticity and trust are fundamental features, not afterthoughts.

To learn more about implementing C2PA with Fastly Image Optimizer or to share your thoughts on content provenance technology, reach out to our team or try Image Optimizer yourself. 

The fight against misinformation is far from over, but with tools like C2PA and the commitment of organizations across the industry, we're building the foundation for a more trustworthy digital future.

Fastly is a proud member of the Coalition for Content Provenance and Authenticity (C2PA). Learn more about C2PA and the technical specifications.