WADL Builder: AI Disclosure

AI use, stated in one line. WADL v1.0: P2-V2-R2 That is a Wandt AI Disclosure Label. WADL is an open standard for disclosing how AI was used in academic and professional work. Three codes answer three questions: what the AI did, how it was checked, and what was kept. Every code has one defined meaning, so a label says the same thing to whoever writes it and whoever reads it. WADL Builder builds one on your device. The standard is published at usewadl.org under a Creative Commons license, CC BY 4.0: anyone may share and adapt it, even commercially, with credit. Version 1.0 of the standard is in beta under active peer review. WADL is a disclosure framework, not an AI detector: a label is the author's own account of how the work was made. WHAT THE STANDARD PROVIDES • One place for it: after your identifying details, before the work itself. • One scope: the whole workflow counts, including AI that shaped the work without appearing in the final version. • One expectation: every level of AI use carries a recommended minimum of checking and records. • One way to set rules: an assignment's AI policy is written in the same codes. • One machine-readable form: a JSON schema, published at usewadl.org. • Two principles: disclosing AI use is not evidence of academic misconduct, and when unsure, disclose and choose the higher process level. HOW A WADL LABEL WORKS WADL v1.0: P2-V2-R2 says AI helped organize and outline the work (P2), it was checked against course materials or reliable secondary sources (V2), and a summary of prompts and outputs was kept (R2). • P, what the AI did: P0 means no generative AI was used; P1 to P5 run from brainstorming up to drafting, rewriting, analysis, coding or generated content. +R, +F and +O add research, feedback or organizing help. • V, how it was checked: from a review for fit and obvious errors (V1) to checking primary or authoritative sources (V3) and documenting corrections (V4). • R, what was kept: from naming the tool and its purpose (R1) to a full prompt-and-output log with verification notes (R4). WADL BUILDER • Build a label: pick what the AI did, name the tools, set the checking and records levels. Say AI was used and the recommended levels fill in; choose less and you get a note, never a block. WADL drafts a narrative for you to edit, then asks you to confirm you remain responsible for the work. • Take it anywhere: copy the label or the full disclosure, or export a PDF record, a PNG badge or a QR code that opens it on usewadl.org. • Set rules in the same terms: the Instructor tab states an assignment's AI rules as an authorization level (AI Encouraged, Permitted, Restricted or Prohibited), the highest process level allowed, and the minimum checking and records. Start from one of eight presets, then copy it as plain text or as Brightspace HTML. • Learn the standard: the Key tab explains every code; Settings adds examples, a user guide and an FAQ. PRIVACY: ZERO TELEMETRY Fully offline: WADL Builder has zero network configurations, makes no network connections of any kind and collects nothing about you. It builds every label, QR code and PDF on your device, reads none of your files, photos or contacts, and has no account or sign-in. Its links to profworkbench.com open in your browser only when you tap them. A disclosure leaves your device only when you copy, share or export it, or someone scans its QR code. Scanning opens usewadl.org with the whole disclosure inside the link. That needs internet, and the site's host, Cloudflare, may keep the link in its request logs. Copied text may reach your other Apple devices through the clipboard. On a shared device, Clear Form empties the Builder; it leaves the clipboard alone. On iPhone, the app switcher shows only the WADL logo, not your form. ABOUT WADL The Wandt AI Disclosure Label was created by Adam Scott Wandt, Associate Professor of Public Policy at John Jay College of Criminal Justice. WADL Builder is made by ProfWorkBench, LLC.