AI image tools embed C2PA metadata, but common processing pipelines silently strip it

·2 views

Major AI image generators including Claude, OpenAI, and Gemini now embed C2PA provenance metadata in their output files to signal content origin and authenticity. However, widely used image processing tools such as Sharp, ImageMagick, and Pillow strip this metadata by default when resizing or converting images, breaking the provenance chain silently. CDN image optimizers and format conversions like JPEG-to-AVIF also typically discard C2PA manifest data unless explicitly configured to preserve it. Developers can detect where metadata is lost by using the c2patool CLI to inspect files at each stage of an upload-and-delivery pipeline. The recommended fix is to selectively strip only sensitive metadata like GPS coordinates rather than removing all metadata blocks wholesale, and to add CI checks that catch provenance loss before it reaches production.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Cursor Launches Origin, a Git Hosting Platform Built for AI-Driven Development

Cursor has announced Origin, its own git hosting and code collaboration platform designed for teams and AI agents. The product covers repository hosting, code review, and collaboration workflows, targeting a gap in the AI coding pipeline that Cursor previously left to services like GitHub and GitLab. Origin is reported to be connected to Cursor's earlier acquisition of Graphite, which specialised in high-volume pull request workflows. The platform is described as built for the speed at which AI agents generate and push code, which Cursor argues exceeds what traditional forges were designed to handle. A public waitlist is now open at cursor.com/origin, with a general release expected this fall; pricing and full documentation have not yet been published.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Hub Sites Don't Fix Microsoft 365 Cross-Team Search — Here's What Actually Does

A common frustration in Microsoft 365 is that files shared across Teams become hard to find, and adding a hub site is often the first attempted fix. However, hub sites only unify navigation and scope search — they do not grant users permission to access content they are not already members of. Adding everyone to every Team restores search visibility but destroys confidentiality boundaries and floods channels with noise, making it an unacceptable trade-off for sensitive projects. The real solution is an information architecture decision: separating active work in member-gated Teams from completed work moved to broadly readable sites, surfaced through a hub. This three-tier approach — active work, a read tier for closed projects, and an exceptions flow via SharePoint access requests — addresses discovery without compromising security.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer releases edfcore, a zero-dependency TypeScript library for parsing EDF biosignal files

A developer has released edfcore, an open-source TypeScript library designed to parse EDF-family biosignal recordings — including EEG, ECG, and sleep-study data — without any external dependencies. The library allows developers to work with biosignal data directly in JavaScript and TypeScript environments such as browsers, Node.js pipelines, and Electron applications. Building the parser revealed several non-obvious complexities in the EDF format, including per-signal sampling rates, annotation data embedded as signal channels, and 24-bit sample values in BDF files that require manual byte reconstruction. The developer noted that subtle parsing errors are especially difficult to detect because they can produce plausible-looking waveforms rather than obvious failures. A companion CLI tool called edf2csv is also available for converting EDF and BDF recordings to CSV format; both projects are MIT licensed and published on GitHub and NPM.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Vector Databases Explained: How They Give AI Apps Memory and Context

Vector databases have emerged as critical infrastructure for modern AI applications, solving a key limitation of large language models — their lack of long-term memory or access to private data. Unlike traditional databases that rely on exact keyword matches, vector databases store high-dimensional numerical embeddings that capture semantic meaning, allowing searches for conceptually similar content even when exact words differ. An embedding model converts unstructured data — text, images, or audio — into arrays of floating-point numbers, where related concepts cluster geometrically close together in multi-dimensional space. These databases power use cases such as Retrieval-Augmented Generation (RAG), which reduces AI hallucinations by fetching relevant context before querying a language model. Popular options range from dedicated services like Pinecone and Weaviate to PostgreSQL extensions like pgvector and embedded solutions like LanceDB, each suited to different scalability and infrastructure needs.

0 comments Read more at DEV Community