The most complete tool for your EU AI Act report
Analyze, detect, document AI, and automate your AI Act compliance in a few clicks.
Fulfai scans your environments, then generates and automates your EU AI Act compliance.
AI systems detected
acme-corp/api-gateway
Complete
acme/platform-ai
prod-ml-bucket
contoso-ai-training
acme-corp/frontend-app
acme-corp/data-pipeline
Complete
Multi-source scanning, 20+ connectors including GitHub
Scan also via the command line
npm install fulfai- ✓No repo OAuth needed — works on local and private code.
- ✓Your source code never leaves your machine — only the AI inventory is sent.
- ✓Drop it into CI/CD to track AI on every build.
Stay permanently ready for a European AI Office audit
Up-to-date AI inventory, risk classification and automatically generated documentation: your compliance file is ready at any time.
Learn more →Understand your obligations and the risks
Regulation (EU) 2024/1689 sets obligations for every company that builds or uses AI. Here's the essential.
Heavy financial penalties
Fines are based on worldwide annual turnover — whichever amount is higher.
Prohibited AI practices (manipulation, social scoring, mass biometric surveillance…)
Breach of the obligations for high-risk systems
Incorrect or misleading information provided to authorities
A risk-based approach
Banned — social scoring, manipulation, real-time mass biometric identification.
Strict obligations — HR, credit, health, biometrics, critical infrastructure.
Transparency — tell users they interact with AI, label generated content.
No obligations — spam filters, games, recommendations.
What your company must put in place
- Inventory of every AI system and agent in use
- Risk classification of each system
- Technical documentation and logging of decisions
- Continuous monitoring and post-deployment tracking
- Human oversight and incident handling
- Transparency toward users and authorities
A timetable already under way
Ban on unacceptable-risk practices
Obligations for general-purpose AI models (GPAI)
Obligations for high-risk systems (Annex III)
AI embedded in regulated products (Annex I)
Fulfai detects your AI systems, classifies them by risk, and automates your documentation and continuous monitoring.
Get started for freeAutomate scans and receive email reports to prove control over your AI system
When a scan finishes, you get an email summarizing the risks and AI systems detected.
Automatic scan report
Fulfai
Repository: acme-corp/api-gateway
Source: github
Date: 12/06/2026 14:32
Risk summary
Detected AI features
Using the OpenAI API for text generation and automated classification.
Version: 1.2.0
AI agent orchestration and prompt chaining over customer data.
Version: 0.3.5
Pre-trained NLP models for sentiment analysis.
Version: 4.41
Email sent automatically by Fulfai
Your whole team on one compliance
Invite your colleagues by email: they join your organization and share the same dashboard, the same AI inventory and the same reports.
Email invitation
The owner adds an employee in one click; they receive a link to create their account and join the organization.
Shared data
Scans, detected AI systems and documents are shared in real time across all members.
Roles and permissions
Assign Admin, Member or Viewer roles to decide who can scan, manage or only view.
A centralized organization
Name your organization and steer the whole company's AI compliance from a single place.
- OwnerSASarah A.sarah@acme.co
- AdminKMKarim M.karim@acme.co
- MemberLDLéa D.lea@acme.co
- ViewerTBTom B.tom@acme.co
Available on the Business plan — unlimited members.
The shortest path to your EU AI Act compliance
Multi-source scanning
GitHub • GitLab • Bitbucket • AWS S3 • Azure Blob • Google Cloud • Microsoft 365 • Databricks • Hugging Face • Apache Airflow • n8n • Slack • ZIP upload
Real-time AI detection
OpenAI, LangChain, TensorFlow, Hugging Face, Anthropic...
Risk classification
Minimal • Limited • High under the AI Act
EU AI Act questionnaire
Complete form with contextual help
Professional PDF report
Compliance score + action checklist
Smart automation
Hourly, weekly, or monthly scans
Customer support that's always within reach
With our built-in ticketing system, your company can contact support at any time — a question about a scan, a report or your compliance, we're here.
- ✓Open a ticket in seconds from your dashboard.
- ✓Email follow-up and replies until full resolution.
- ✓Priority support with a 24-hour response for the Business plan.
6 steps to stay compliant
Analyze your sources
Connect GitHub, GitLab, AWS S3, Azure Blob repositories or upload your files and folders.
Automatic risk detection
Our engine identifies AI libraries and classifies them by risk level (Minimal / Limited / High).
Complete the EU AI Act questionnaire
Answer mandatory questions with contextual help (DPIA, governance, transparency, bias...).
Generate your PDF report
Get a professional report with compliance score, detected systems list, and action checklist.
Plan your next actions
Follow precise recommendations to reduce risks and meet regulatory requirements.
Automate compliance
Schedule recurring scans (hourly, every Monday, or on the 1st of each month).
Fill your documents 10× faster
AI drafts every field of your AI Act, GDPR, DPIA records. You approve, it fills.
Déclaration de conformité RGPD
Registre des traitements · Art. 30
No data is stored
We never store your source code. Our scanner only looks for AI agents and AI libraries inside your dependencies — nothing else is kept.
- ✓Your source code is never stored
- ✓Only AI agents and AI libraries are searched
- ✓Aligned with GDPR data minimisation
What we scan
We analyse dependency manifests and import statements (for example package.json, requirements.txt, lockfiles) to identify AI libraries and AI agent frameworks. The scan looks for known AI providers and packages — not the content of your files.
What we do NOT store
We do not retain your source code, repository contents, personal data, or proprietary logic. Only the resulting inventory of detected AI agents and AI libraries (name, version, risk classification) is saved to your account.
How the scan works
The scan reads dependency metadata in memory during the analysis. Once the AI inventory is produced, the raw code is discarded. You can delete any scan and its results at any time from your dashboard.
You receive a personalized data non-retention certificate: this document commits us to analysing your environments without collecting any information other than AI libraries and agents.
Ready to secure your AI compliance?
Join teams anticipating European regulation.
Get started for freeNo credit card • Try it for free