The Horizon Europe programme has launched a call for research to develop robust, safe and reliable AI solutions, particularly generative ones, that operate in real-world conditions. These solutions should provide adequate, meaningful and complete explanations when relevant.

The research should enhance trust in AI solutions and increase transparency while leading to an acceptable loss of accuracy and efficiency. Ideal solutions will apply to foundational models and machine learning.

Activities are expected to start at TRL 2-3 and achieve TRL 4-5 by the end of the project.

Research Directions

• Creating Stronger Algorithms to train AI to resist attacks and ensure reliability
• Making AI Understandable through new methods and tools that explain how AI models work
• Predicting Limits of AI by developing algorithms that monitor and predict AI behaviour
• Balancing Performance of AI to enhance AI robustness, explainability and efficiency
• Ensuring Reproducibility using standard benchmarks and evaluation methods
• Foundational research focused on AI and machine learning

Eligibility Criteria

Entities legally established in:

• EU Member States, its Overseas Countries and Territories (OCT) or Horizon Europe associated countries

Funding

• Up to €7,500,000

Please follow this link for more information on the scope of the call and the procedural requirements for the applications.

The deadline for application is 18 September 2024, 17:00 CEST (Brussels time).

Do you have any questions about this call? Do you want to improve your proposal? Are you looking for project partners? Join our interactive webinar and matchmaking session on 12 July 2024 to gain in-depth insights about the call. Registration is now open at this link.