AI Act Article 9 Explained: Data and Data Governance Requirements for US and EU Companies
What are the data governance requirements under AI Act Article 9? Learn how to ensure data quality, documentation, and compliance for high-risk AI systems. Avoid fines with this step-by-step guide.
AI Act Article 9 Explained Data and Data Governance Requirements for US and EU Companies
Last updated May 17 2026 Reading time 10 minutes
Good Data Is the Foundation of Compliant AI
The EU AI Act does not just regulate AI systems. It regulates the data that powers them. Article 9 sets the standards for data governance. These standards apply to high risk AI systems.
Poor data leads to poor decisions. Biased data leads to biased outcomes. Inaccurate data leads to harm. The EU wants to prevent this.
For US companies this is a critical point. If your AI system is used in the EU and falls under the high risk category you must comply with Article 9. The requirements are strict. The penalties for non compliance are severe.
This guide will explain what Article 9 requires. It will show you how to ensure data quality. It will walk you through the governance obligations. It will help you build a compliant AI system.
What Is Article 9 of the EU AI Act
Article 9 is about data. It sets the requirements for data used in high risk AI systems. These requirements ensure that the data is
Relevant Representative Free from errors Complete Up to date
The goal is to prevent harm. The goal is to ensure fairness. The goal is to build trust in AI systems.
Why Data Governance Matters
Data is the fuel of AI. Without good data AI systems cannot function properly. They cannot make fair decisions. They cannot avoid bias.
Here is why data governance is crucial.
Prevents Bias
Biased data leads to biased AI. This can discriminate against certain groups. It can violate fundamental rights. Good data governance helps prevent this.
Ensures Accuracy
Inaccurate data leads to wrong decisions. In healthcare this can be life threatening. In finance this can cause significant harm. Good data governance ensures accuracy.
Builds Trust
Trust is essential for AI adoption. Users must trust that AI systems are fair. They must trust that AI systems are reliable. Good data governance builds this trust.
Complies with Regulations
The EU AI Act requires good data governance. But so do other regulations. GDPR requires data protection. Sector specific laws may have additional requirements. Good data governance ensures compliance.
The Data Governance Requirements of Article 9
Article 9 sets specific requirements for data governance. Here is what you need to know.
Data Quality Assurance
You must ensure the data used to train test and validate your AI system is of high quality. This means
The data must be relevant to the AI system’s intended purpose The data must be representative of the real world scenarios the AI system will encounter The data must be free from errors and biases The data must be complete and up to date
Data Documentation
You must document the data used in your AI system. This includes
The sources of the data The collection methods The preprocessing steps The characteristics of the data
Data Traceability
You must ensure that the data used in your AI system is traceable. This means you must be able to track
Where the data came from How the data was processed How the data was used in the AI system
Data Retention
You must retain the data and documentation for a specified period. This allows for audits. This allows for investigations. This ensures accountability.
Data Security
You must ensure the data is secure. This includes protecting the data from
Unauthorized access Data breaches Data loss
How to Ensure Data Quality
Ensuring data quality is a multi step process. Here is how to do it.
Step 1 Define Data Requirements
Start by defining what data you need. What is the intended purpose of your AI system? What data is relevant to this purpose?
Example An AI system for medical diagnosis needs high quality medical images. It needs diverse patient data. It needs accurate labels.
Step 2 Source High Quality Data
Collect data from reliable sources. Ensure the data is representative. Ensure it covers all relevant scenarios.
Example For a credit scoring AI use data from diverse financial institutions. Ensure it covers all demographic groups. Ensure it is free from biases.
Step 3 Clean and Preprocess Data
Clean the data to remove errors. Preprocess the data to ensure it is in the right format. This includes
Removing duplicates Correcting errors Normalizing data Handling missing values
Step 4 Validate Data
Validate the data to ensure it meets your requirements. This includes
Checking for biases Testing for accuracy Ensuring completeness
Step 5 Document Data
Document all aspects of the data. This includes
The sources of the data The collection methods The preprocessing steps The characteristics of the data
Real World Examples of Data Governance
Healthcare AI
A US based healthcare AI company develops a diagnostic tool. The tool uses medical images to diagnose diseases. The company must ensure the data is of high quality.
This includes
Using diverse and representative medical images Ensuring accurate labels for the images Documenting the sources and preprocessing steps Validating the data for biases and errors
HR AI
A multinational company uses an AI system for recruitment. The system screens job applicants. The company must ensure the data used is of high quality.
This includes
Using diverse and representative candidate data Ensuring the data is free from biases Documenting the data sources and preprocessing steps Validating the data for accuracy and completeness
Credit Scoring AI
A US fintech company develops an AI system for credit scoring. The system assesses the creditworthiness of loan applicants. The company must ensure the data is of high quality.
This includes
Using diverse and representative financial data Ensuring the data is free from biases Documenting the data sources and preprocessing steps Validating the data for accuracy and completeness
Common Mistakes in Data Governance
Mistake 1 Using Low Quality Data
Low quality data leads to poor AI performance. It can lead to biased outcomes. It can lead to harm.
Always use high quality data. Always validate the data. Always document the data.
Mistake 2 Ignoring Data Bias
Biased data leads to biased AI. This can discriminate against certain groups. It can violate fundamental rights.
Always check for biases. Always use diverse and representative data. Always validate the data for fairness.
Mistake 3 Failing to Document Data
Documentation is crucial for compliance. It is also crucial for audits. It is crucial for investigations.
Always document the data. Always document the sources. Always document the preprocessing steps.
Mistake 4 Not Ensuring Data Security
Data security is essential. Data breaches can lead to harm. They can lead to legal consequences.
Always ensure data security. Always protect the data from unauthorized access. Always protect the data from breaches.
Mistake 5 Not Retaining Data
Data retention is required for compliance. It is also required for audits. It is required for investigations.
Always retain the data. Always retain the documentation. Always ensure traceability.
How DilAIg Helps with Data Governance
Ensuring data governance can be complex. DilAIg simplifies the process.
Our tool guides you through each step. It helps you define data requirements. It assists in sourcing high quality data. It ensures data quality and documentation.
For US companies our tool ensures compliance with both US and EU regulations. It flags EU specific requirements. It helps you navigate the complexities of the AI Act.
Here is how it works.
1 Answer a series of questions about your AI system. What data does it use? Where does the data come from? How is the data processed?
2 Our tool analyzes your responses. It identifies potential data quality issues. It flags biases and errors.
3 We generate a comprehensive data governance plan. It includes all the necessary steps. It is ready for implementation.
4 We provide guidance on data documentation. We help you ensure data traceability. We assist with data security.
Ensure your data governance complies with Article 9. Start Your Data Governance Check
FAQ Data Governance Requirements
Q What is Article 9 of the EU AI Act
Article 9 sets the data governance requirements for high risk AI systems. It ensures that the data used is of high quality representative and free from biases.
Q Why is data governance important for AI systems
Data governance ensures that AI systems are fair accurate and reliable. It prevents harm. It builds trust. It ensures compliance.
Q What are the data quality requirements under Article 9
The data must be relevant representative free from errors complete and up to date.
Q How can I ensure data quality for my AI system
Define data requirements. Source high quality data. Clean and preprocess the data. Validate the data. Document the data.
Q What are the documentation requirements for data
You must document the sources of the data the collection methods the preprocessing steps and the characteristics of the data.
Q What happens if I do not comply with Article 9
You could face fines up to 7% of your global revenue or 35 million euros. You could also face reputational damage and loss of trust.
Key Takeaways
Article 9 of the EU AI Act sets data governance requirements for high risk AI systems. Good data governance ensures that AI systems are fair accurate and reliable.
Data quality is crucial. Ensure the data is relevant representative free from errors complete and up to date. Documentation is also crucial. Document the sources collection methods preprocessing steps and characteristics of the data.
Common mistakes include using low quality data and ignoring data bias. DilAIg’s tool simplifies data governance. It guides you through each step. It generates the necessary documentation.
Next Steps
Ensure your data governance complies with Article 9. Start Your Data Governance Check
Implement data quality assurance. Learn How
Need help? Book a Demo
Join the Conversation
How do you ensure data quality for your AI systems? What challenges have you faced with data governance? Share your thoughts in the comments or tweet us @DilAIg.
Further Reading
Official EU AI Act Text Article 9 European Commission Data Governance Guidelines DilAIg’s AI Act Compliance Hub Best Practices for AI Data Quality