New standards available in Foxit Editor

Foxit Editor 2026.2 brings two standards-based capabilities to market: content provenance verification via C2PA https://c2pa.org/, and AI-assisted formula accessibility via MathML.

With the 2026.2 release, Foxit Editor adds support for two open standards that address different but related problems in modern document workflows — knowing whether visual content can be trusted and making mathematical content genuinely accessible. Together, they mark a step toward PDFs that carry richer, machine-readable context about what they contain and where it came from.

C2PA: seeing where an image came from

AI generation and AI-assisted editing of images have become part of everyday content production, from marketing photography to e-commerce listings. That shift has made it harder for anyone viewing an image to know whether it is an untouched photograph, a heavily edited version of one, or something generated entirely by a model — a problem that becomes especially acute inside PDFs, where images are routinely embedded in contracts, financial reports, compliance documents, and other content meant to be taken at face value.

The Coalition for Content Provenance and Authenticity (C2PA) addresses this with an open technical standard for attaching a tamper-evident “manifest” to digital content — a record of its origin, the tools that touched it, and the edits made along the way. If content is altered after the manifest is attached, that tampering becomes detectable.

In Foxit Editor 2026.2, when an image carrying a valid C2PA manifest is placed into a PDF, Editor automatically detects it and displays a C2PA badge directly on the image — no manual verification step required. Anyone reviewing the document can see at a glance that the image carries verifiable provenance data, with the signal traveling alongside the content itself rather than living in a separate report or tool.

This first release covers JPG images with an intact manifest; the manifest has to survive the trip from wherever the image was generated or edited into the PDF, which depends on upstream tools preserving rather than stripping it. Support for additional image formats and for other embedded object types is planned for future releases, extending provenance visibility to more of what a PDF can contain.

MathML: from reading accessible math PDFs to creating them

After adding support for reading PDF files that already contain MathML, Foxit is now extending that standards-based accessibility work into creation. Foxit Editor 2026.2 does not only consume MathML-enabled PDFs; it also adds tools for producing them by converting detected formulas into structured MathML that can be associated directly with the document’s tag structure. This moves the workflow from passive support to active remediation: teams can take existing math-heavy PDFs and help turn formulas that were previously only visible into formulas that assistive technologies can understand.

Formula to MathML: making equations readable, not just visible

Mathematical formulas in PDFs have historically been rendered as images: visually precise, but invisible to a screen reader. That leaves students, researchers, and professionals who rely on assistive technology without access to the underlying structure of the math itself — a real barrier to STEM content.

Foxit Editor 2026.2 introduces Formula to MathML, an AI-powered feature, found under the Accessibility tab, that detects formulas in a tagged PDF and converts them into MathML, the W3C standard for representing mathematical notation in a structured, searchable, and accessible form. Foxit has also published a Formula to MathML overview describing the workflow.

The feature requires a tagged PDF as a starting point — it completes accessibility remediation rather than replacing the initial tagging step. From there, Editor scans the document to build a list of detected formulas; users select one or more and convert them with a single action. The dialog shows the original formula alongside the AI-generated preview for visual verification, and the resulting MathML is presented as editable source for anyone who wants to inspect or fine-tune it before committing. Once approved, the MathML is inserted as an associated file directly into the formula's node in the structure tree, making the formula readable by assistive technology from that point on. Conversion can be applied selectively, formula by formula, rather than as an all-or-nothing pass across the whole document.

This directly supports compliance with accessibility standards such as PDF/UA and WCAG, which increasingly expect mathematical content to be machine-readable rather than purely visual. It is aimed at publishers and content creators producing STEM material, educational institutions preparing accessible course materials, engineering and scientific teams documenting formulas in reports and specs, and accessibility and compliance officers responsible for meeting these requirements. The broader standards context is captured in the PDF Association’s Best Practice Guide: Math in PDF and in Foxit’s earlier PDF Association article, Accessible math – Foxit is spearheading the implementation of new PDF features. Beyond accessibility, MathML-tagged formulas are also searchable, indexable, and easier to reuse in other tools.

MathML related links you may want to open:

• https://pdfa.org/accessible-math-in-pdf-finally/
• https://pdfa.org/accessible-math-foxit-is-spearheading-the-implementation-of-new-pdf-features/
• https://pdfa.org/pdf-association-announces-publication-of-new-best-practice-guide-for-mathematical-content-in-pdf/

A shared direction

Both features point in the same direction: PDFs that expose more of what they contain — provenance for visual content, structure for mathematical content — in ways that standards bodies and assistive technologies can act on directly, rather than leaving that context locked inside a flat, visual-only page. Both are first steps rather than finished coverage, with broader format and object-type support on the roadmap for C2PA, and continued refinement expected for formula detection and conversion.

Foxit Editor 2026.2 is available now, https://www.foxit.com/pdf-editor/.