EU model contractual AI clauses to pilot in procurements of AI

← All articles
Artificial intelligence · EN

EU model contractual AI clauses to pilot in procurements of AI

The EU has introduced new model contractual AI clauses, a significant step towards fostering responsible and trustworthy AI development.

Starting my legal journey with procurement, I vividly remember the importance of data protection when procuring IT software. This ignited my hunger to delve deeper into this crucial topic, leading me to specialize in data protection. Now, I’m thrilled to share the latest developments in this field.

Public organizations seeking to procure AI systems from external suppliers can voluntarily use these standard EU model contractual AI clauses. As the proposed AI Act undergoes negotiations, organizations can assess on a case-by-case basis whether these clauses are suitable for their specific AI system procurement.

🔒 High Risk and Non-High Risk Versions:
The High Risk version targets AI systems classified as ‘high-risk’ under the AI Act (check Annex III). For non-high risk AI, application is not mandatory but recommended to enhance trustworthiness. The Non-High Risk version is designed for such AI systems.

🌍 Extended Application:
Public organizations, based on impact, may extend these clauses to other algorithmic systems beyond AI, ensuring increased accountability, control, and transparency.

📜 Key Considerations:
The clauses are specific to AI systems and proposed AI Act matters.
They do not cover obligations under other legislation (e.g., GDPR).
Customization is essential, addressing specifics like intellectual property, acceptance, payment, delivery times, applicable law, or liability.

🗣️ Your Next Steps:
Carefully read and customize these clauses to your organization’s context and procurement needs. They’re available in all EU languages: https://lnkd.in/epSi-VQH.

#AI #EURegulations

This post has a document attached on LinkedIn. View on LinkedIn →

Discuss on LinkedIn