Skip to content

Latest commit

 

History

7 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 

Repository files navigation

Title GPAI Code of Practice Self-Assessment Tool
ColorFrom Blue
ColorTo Indigo
Sdk Streamlit
Sdk_version 1.35.0
App_file app.py
Pinned true
License MIT
Short_description Check your GPAI model compliance with EU AI Act Articles 53 and 55

GPAI Code of Practice Self-Assessment Tool

A free, open-source self-assessment tool for providers of general-purpose AI (GPAI) models to check their compliance with the EU AI Act GPAI Code of Practice obligations under Articles 53 and 55.

Built by: Suhana Sayyad | MSc Cybersecurity, TUS Athlone

GitHub: github.com/SuhanaSayyad


What this tool does

The GPAI Code of Practice was finalised on 10 July 2025. It sets out the concrete compliance measures that GPAI model providers must implement under Articles 53 and 55 of the EU AI Act. These obligations became applicable on 2 August 2025, with Commission enforcement powers beginning 2 August 2026.

This tool walks through 13 key measures across three chapters and generates a full compliance assessment with downloadable reports.

Measures covered:

  • Transparency (T1-T5) - Technical documentation, training data summary, downstream provider information, instructions for use, and AI content marking. Applies to all GPAI providers.
  • Copyright (C1-C3) - Copyright compliance policy, opt-out implementation, and training data documentation. Applies to all GPAI providers.
  • Systemic Risk (SR1-SR5) - Adversarial testing, systemic risk assessment, risk mitigation, incident reporting, and cybersecurity protection. Applies only to models trained above 10^25 FLOPs or designated by the EU AI Office.

Features

Assessment

  • 13 measures mapped to specific EU AI Act articles
  • Four compliance statuses per measure: Fully implemented / Partially implemented / Not implemented / Not applicable
  • Implementation guidance for every measure explaining what is required
  • Evidence checklist per measure listing the exact documents an auditor would ask for
  • Notes field per measure to record your implementation evidence and references

Results

  • Compliance score with a red/amber/green (RAG) colour-coded banner (green above 80%, amber above 50%, red below 50%)
  • Radar chart showing compliance percentage across Transparency, Copyright, and Systemic Risk sections
  • Colour-coded result card for every measure showing status and your notes
  • Automatic skip of Systemic Risk section for models below the threshold

Action plan

  • 90-day action plan automatically generated from your gaps
  • Gaps categorised by implementation complexity: 30 days for quick wins, 60 days for medium effort, 90 days for complex implementation
  • Priority categories based on actual effort required, not just article order

Reports

  • Download as PDF: professional formatted report with score banner, all measure results, notes, and action plan
  • Download as TXT: plain text version for import into compliance management systems
  • Shareable URL: browser URL automatically updates with your answers so you can bookmark or share your exact assessment state

Who this is for

  • Providers of foundation models, large language models, and multimodal models placing systems on the EU market
  • Compliance teams and DPOs at AI companies assessing GPAI obligations
  • Legal and governance teams advising on EU AI Act implementation
  • Researchers and practitioners studying the GPAI Code of Practice

Legal basis

Measure ID EU AI Act Reference
Technical documentation T1 Article 53(1)(a) + Annex XI
Training data summary T2 Article 53(1)(d)
Downstream provider info T3 Article 53(1)(b)
Instructions for use T4 Article 53(1)(b)
AI content marking T5 Article 50(5)
Copyright compliance policy C1 Article 53(1)(c)
Opt-out implementation C2 Article 53(1)(c)
Training data documentation C3 Article 53(1)(c)
Adversarial testing SR1 Article 55(1)(a)
Systemic risk assessment SR2 Article 55(1)(b)
Risk mitigation SR3 Article 55(1)(b)
Incident reporting SR4 Article 55(1)(c)
Cybersecurity SR5 Article 55(1)(d)

How to use

  1. Click Start Assessment
  2. Enter your model name and provider details
  3. Confirm whether your model is a GPAI model and whether it has systemic risk designation (above 10^25 FLOPs)
  4. Work through each section answering the compliance status for every measure
  5. Add implementation notes and evidence references in the notes field under each measure
  6. View your results including the radar chart and 90-day action plan
  7. Download your PDF or TXT report

The assessment takes approximately 10 to 15 minutes to complete. Your answers are encoded in the URL as you go so you can bookmark or share at any point.


Technical stack

Component Library
Web framework Streamlit 1.35.0
PDF generation reportlab 4.1.0
Radar chart Plotly 5.22.0
Language Python 3.10+
Hosting Hugging Face Spaces

Running locally

git clone https://github.com/SuhanaSayyad/gpai-cop-self-assessment.git
cd gpai-cop-self-assessment
pip install -r requirements.txt
streamlit run app.py

Disclaimer

This tool provides technical guidance only and does not constitute legal advice. It does not replace qualified legal counsel or formal conformity assessment. Always verify your obligations against the official EU AI Act text at artificialintelligenceact.eu.

Assessment content is based on analysis of the GPAI Code of Practice and EU AI Act text available as of June 2026.


Related tools

Tool Description
EU AI Act / NIST AI RMF / ISO 42001 Crosswalk Maps 30 EU AI Act obligations to NIST and ISO 42001 with strength indicators and gap analysis
GPAI CoP Self-Assessment This tool

Built by Suhana Sayyad | GitHub | MSc Cybersecurity, TUS Athlone, Ireland

About

Free open-source self-assessment tool for GPAI model providers to check compliance with EU AI Act Articles 53 and 55. 13 measures, radar chart, PDF report, 90-day action plan, evidence checklists. Built with Python and Streamlit.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages