Vision

CERTAIN seeks to create a unified and compliant ecosystem for AI stakeholders, fostering trust, transparency, and innovation in the European data economy. The project's main goal is to help organizations navigate complex regulations, making them aware of the benefits that come with advanced technology as well as encourage them to make use of these technologies. Enabling sustainable growth in the data and AI sectors through collaboration and standardization is also a core focus of the project.

Goals

CERTAIN aims to set up a framework that facilitates responsible AI development, raises the bar for data governance, and makes the societal benefits of AI innovation tangible for a wide range of stakeholders. This requires completing the sub-goals listed below:

• Make it possible to trace critical information of AI systems.
• Create guidelines for assessing AI system that comply with ethical standards and (EU) regulations.
• Design tools for dataspace providers and data holders to help them meet EU AI regulations and minimize energy consumption.
• Design certification procedures for AI systems.
• Collect empirical evidence on the applicability and effectiveness of the proposed framework across multiple sectors.
• Set the stage for the development of an open, dynamic, multi-disciplinary community centred on EU AI ecosystems.

Methods

Overall, our project can be divided into a few key activities. First, we engage in an intensive exchange with various stakeholders to familiarize ourselves with their needs. We identify the primary requirements, and assign them to four categories: legal, ethical, technical, and certification. Second, we organize so-called living labs, where researchers, businesses, and users come together to develop and test new ideas, products, or services in real-life testing environments. During this stage of the project, we build prototypes and guidelines and go through multiple loops of gathering feedback and refining the product. Such an iterative process guarantees that stakeholder needs are thoroughly considered. Finally, we validate the tools, guidelines, and certification process by testing them in an operational environment.        

Outcome and Impact

In the CERTAIN project we collect and piece together information of the regulatory landscape surrounding AI (e.g., EU AI Act). We bring attention to the problems and the pitfalls and come up with a comprehensive and user-friendly solution that supports businesses and the public sector in navigating the complexities of the regulations. At the same time, we encourage innovation and uphold ethical AI practices. Throughout the project we pursue a multi-faceted approach and gather inputs from a wide range of stakeholders, partners, legal experts, and AI providers.

Together with them we streamline AI certification, thereby lowering the time-to-market and development costs.

Overall, the CERTAIN project promotes open and trusted European data spaces by establishing standards for data interoperability and compliance. It also encourages transparent, responsible and privacy-preserving data solutions and places – in accordance with the EU's Green Deal – high value on energy-efficient AI and sustainable data methods.