- Anh Quan Tran
- Van Loc Nguyen
- Nguyen Khanh Toan Tran
- Viet Tri Ong
- Khoi Nguyen Bui
AIMS Hackathon – AI-powered Modern Slavery Reporting & Benchmarking
ChainBreaker is a web platform that analyses modern slavery statements, surfaces strengths and weaknesses, produces actionable reports, and benchmarks organisations across peers and time.
The goal is to move reporting from manual, fragmented, and compliance-minimal to comparable, auditable, and impact-oriented.
Modern slavery affects more than 50 million people worldwide, often hidden in global supply chains (International Labour Organization (ILO), Walk Free & International Organization for Migration (IOM) 2022).
Governments require organisations to publish modern slavery statements outlining how they identify and mitigate risks.
In Australia, only 14% of companies report finding cases of modern slavery in their supply chains, despite global estimates suggesting much higher prevalence (PASA 2025).
But today’s reporting ecosystem is flawed:
-
Inconsistent formats & low comparability
Statements appear as PDF, DOCX, or HTML, with divergent structures, making cross-company or cross-year analysis unreliable. -
Manual, fragmented workflows
Compliance teams spend weeks compiling information from scattered spreadsheets, reports, and websites. -
Box-ticking incentives
Many companies produce vague, narrative-driven statements that avoid measurable KPIs, third-party assurance, or supply-chain depth. -
Weak feedback loops
With few standardised metrics and limited enforcement, NGOs and investors struggle to hold organizations accountable.
Result: Stakeholders cannot compare like-for-like, prioritize risks, or track progress credibly. Compliance remains a burden rather than a driver of change.
- Automate ingestion and analysis of statements across formats
- Standardise disclosures into a scoring & evidence schema
- Provide interactive dashboards (KPI trends, gaps, peer benchmarks) and exportable reports
- Enable auditable AI: every claim links back to verbatim evidence
- Reduce the reporting burden while raising the bar on quality and comparability
ChainBreaker is built as a full-stack web platform (Frontend React/Vite, Backend Node/Express).
-
Ingestion & Normalization
Upload PDFs/links → parse → segment → normalise into a disclosure schema (policy, due diligence, risk assessment, remediation, training, KPIs, governance, supply-chain mapping). -
AI-Assisted Evidence Extraction
Extracts claims and anchors them to citations (page/section). Flags ambiguity, vagueness, and unsupported assertions. -
Scoring & Benchmarking
Transparent rubric with weighted criteria and explain-why rationales. Enables comparison across peers, sectors, and years. -
Reports & Dashboards
Instant Strengths/Weaknesses summaries, Gap Analysis, KPI Trends, Peer Positioning. Export to PDF/CSV. -
Governance & Auditability
Every score is backed by explanation + evidence. Human-in-the-loop review process for verification.
A short pitch video presenting the problem and our solution:
- 100 MB per file
- 10 GB per repository
Datasets are stored in the /datasets folder.
Includes:
- Public modern slavery statements (PDF/HTML)
- API access to compliance and NGO registers
- Notes on data quality issues and transformations
- Containers for merged or generated datasets
Project code is located in the /project folder.
It includes:
- Data transformations, merging, and quality assurance scripts
- AI/ML model code for evidence extraction and scoring
- Frontend (React/Vite) and backend (Node/Express) application code
Supporting docs are stored in the /docs folder.
This may include:
- Presentation slides
- Flyers and promotional materials
- Demo videos
- Protocols and technical guides
All work is created as part of the AIMS Hackathon for research, educational, and demonstrative purposes.
Please refer to the LICENSE file for terms of use and intellectual property rights.
We would like to acknowledge:
- AIMS Hackathon organizers and mentors for providing the platform, guidance, and feedback that shaped this project.
- NGOs, researchers, and government registries working tirelessly to make modern slavery data transparent and accessible.
- Our fellow hackathon participants and peers, whose ideas and collaboration inspired us throughout the event.
- Open-source communities whose tools and libraries (React, Vite, Node.js, Express, AI/ML frameworks) made rapid prototyping possible.
- Early testers and advisors who provided feedback on our scoring rubric, data quality, and dashboards.
ChainBreaker is built on the collective efforts of a community committed to turning compliance into real accountability.
International Labour Organization (ILO), Walk Free & International Organization for Migration (IOM) 2022, Global Estimates of Modern Slavery: Forced Labour and Forced Marriage, International Labour Organization, viewed 18 September 2025, https://www.ilo.org/publications/major-publications/global-estimates-modern-slavery-forced-labour-and-forced-marriage.
PASA 2025, Australia’s modern slavery laws are failing, according to new research - PASA, PASA, viewed 18 September 2025, https://procurementandsupply.com/australias-modern-slavery-laws-are-failing-according-to-new-research/.
This project builds on the open research of Project AIMS (AI against Modern Slavery) by Mila and QUT.
GitHub repository: ai4h_aims-au.
- Describe here the resources used in developing your solution (e.g. GPUs, etc).
This repository and its accompanying models, datasets, metrics, dashboards, and comparative analyses are provided strictly for research and demonstration purposes.
Any comparisons, rankings, or assessments of companies or organizations are exploratory in nature. They may be affected by incomplete data, modeling limitations, or methodological choices. These results must not be used to make factual, legal, or reputational claims about any entity without independent expert review and validation.
Do not use this repository’s contents to make public statements or claims about specific companies, organizations, or individuals.
By submitting this solution to the AIMS Hackathon, our team acknowledges and agrees to abide by the Event’s Terms and Conditions.