Know the Source. Own the Story.
InceptID is an open-source AI provenance and authenticity framework designed to bring traceability and accountability to digital media in a world where content is constantly edited, remixed, and redistributed across platforms. It establishes a unified system for identifying the original source of images, videos, audio, and text, even when metadata has been stripped or content has been heavily transformed. At its core, InceptID aims to rebuild trust in digital information by making origin, authorship, and modification history visible and verifiable.
The system combines multimodal AI fingerprinting with cross-platform intelligence to track how content moves through the internet. It analyzes perceptual features, semantic embeddings, compression artifacts, and structural patterns to identify matches between versions of the same content. By correlating these signals across platforms, InceptID constructs a detailed lineage graph that shows where content originated, how it evolved, and how it spread over time.
A key feature of InceptID is its embedded steganographic provenance layer, which stores invisible but durable metadata directly inside media files. This includes creator attribution, creation timestamps, AI model or tool information, and workflow details such as prompts and editing pipelines. This embedded layer is designed to persist through compression, resizing, screenshots, and format conversion, ensuring that provenance data travels with the content itself.
In addition to origin tracking, InceptID includes advanced deepfake and manipulation detection systems. These modules analyze visual, audio, and textual inconsistencies to identify synthetic or altered media, reconstruct edit histories, and assign confidence scores for authenticity. This makes it possible not only to trace where content came from, but also to understand whether and how it has been manipulated.
Security and anti-abuse systems further strengthen the framework by detecting impersonation, attribution spoofing, coordinated misinformation campaigns, and automated manipulation networks. Combined with cryptographic verification, transparency logs, and decentralized node support, InceptID creates a resilient trust layer for digital media. Its modular, AGPL-licensed architecture ensures that these capabilities remain open, extensible, and resistant to centralized control.
- InceptID is an open-source AI provenance and authenticity framework that traces digital media back to its original source and verifies its transformation history across platforms.
The first public announcement of the InceptID project outline was made on February 25, 2026, on Substack. That initial post introduced the core vision of building an open-source provenance framework capable of tracing digital media back to its original source across platforms. It outlined the foundational ideas behind cross-platform tracking, multimodal fingerprinting, and embedded provenance, setting the stage for InceptID’s development into a full modular system for media authenticity and manipulation detection.