Join CactusSoft in Friendly, MD (onsite) and work on practical software capabilities that emphasize responsible engineering, interoperability, and accessibility. This role focuses on building detection and watermark-related functionality with clear boundaries, transparent disclosure, and standards-driven implementation.
How the work is expected to be delivered
The framework in practice outlines specific engineering outcomes you will help implement across multi-layer marking, detection availability, interoperability, and compliance behavior for AI-related content.
- Multi-layer marking: apply two layers consisting of signed and timestamped metadata, plus an imperceptible watermark embedded in the content itself.
- Detection as an obligation: make a detection solution available free of charge as one of the following: a public specification, distributable software, or an API, and open it to the public.
- Limit of text size: content under 200 tokens falls outside the watermarking obligation.
- Binding interoperability: implement an interoperability solution for detection mechanisms by 2 February 2027.
- Disclosure specifications: use the label consisting of the capitalised acronym “AI”, and provide a freely available EU icon for both “AI GENERATED” and “AI MODIFIED” variants.
- Accessibility rules are mandatory: align with the European Accessibility Act, WCAG 2.1 Level AA, ensure screen-reader detectability, and include haptic cues for audio.
- Labelling boundary with human editorial responsibility: AI-generated published text falls outside the labelling duty once it has been through human review and an identified natural or legal person assumes editorial responsibility, with contact details published (Article 50(4)).
What you’ll be working on
You will help design and implement Node.js components that support watermarking and detection behaviors, enforce the defined content thresholds, provide public-facing detection interfaces, and deliver interoperability for detection mechanisms by the specified deadline.
Core compliance and UX considerations
Engineering choices must follow the required accessibility and disclosure rules. This includes ensuring detectability by assistive technologies and applying consistent labeling behavior using the specified “AI” acronym and the EU icon variants, while respecting the stated exclusions and Article 50(4) conditions.
Location
Friendly, MD with onsite expectations.
