EU Frontier AI Initiative: Developing frontier AI solutions that are safe and computationally efficient within Apply AI (RIA)

Expected Outcome:

The Apply AI Strategy [1] also seeks to bolster EU capabilities and achieve excellence in AI to support the development of European frontier models. As part of the Frontier AI Initiative, which brings together Europe’s leading actors in the field, this topic will support the development of sovereign frontier AI ensuring safety by design. This topic directly contributes to the Apply AI Strategy. Project results are expected to contribute to all of the following expected outcomes:

• Strengthened European capabilities in the development of frontier AI models.
• Improved computational efficiency of frontier AI models, resulting in reduced computational costs.
• Enhanced safety of advanced AI systems based on frontier AI models through the development and implementation of safe-by-design principles and/or AI agents acting as safety evaluators.

Scope:

To advance developments of frontier AI models towards highest-level performance, while ensuring energy efficiency, addressing computational constraints, and strengthening safety. The approach of this topic is twofold. First, it aims to advance the AI field through the development and training of a frontier AI model. The AI model should demonstrate state-of-the-art performance, have multimodal capabilities, and be optimized for agentic AI capabilities such as tool use, reasoning, and autonomous problem-solving. Second, this topic supports research on comprehensive methods to reduce the computational demands of frontier AI models and to ensure their safety, including technical methodologies such as automated testing and interpretability. The primary drivers behind computational efficient AI systems are the urgent challenges posed by the growing energy footprint of AI and current computational limitations. Modern AI models, especially frontier AI models, require substantial computational resources, with a significant impact in the environment. Additionally, they create barriers to entry to those interested in advancing the AI field. Key research areas include compression and distillation techniques aimed at reducing the complexity of large AI models. Innovations in AI architectures are also relevant, with a focus on innovative models that significantly lower computational demands for training and inference. Further, algorithmic approaches aimed at minimizing computational load during pre-training, post-training, and inference can also be considered. Ensuring the safety of AI systems is essential, especially as AI models become increasingly sophisticated and pervasive. Potential research areas to be considered include addressing misalignment, particularly the unintentional misalignment of large AI models. Work in this area could explore methods to detect and mitigate sophisticated misbehaviour, such as alignment faking, reward hacking of human oversight, and encoded reasoning in chain-of-thought (CoT). Additionally, research could focus on enhancing robustness against adversarial attacks, jailbreaks, and backdoors. Further potential areas for innovation include advancing AI models transparency and interpretability. Safety research could also consider risks that may arise when embedding frontier models within agentic AI frameworks, significantly contributing to the trust and safe adoption of powerful AI solutions. This topic contributes to the EU Frontier AI initiative. The project should establish strong links with the Resource for AI Science in Europe (RAISE), ensuring that its priorities inform the research topics addressed. Activities are expected to involve the European AI research community and attract and retain top AI talent working on frontier models and related areas. All proposals are expected to incorporate mechanisms for assessing and demonstrating progress, including qualitative and quantitative KPIs, benchmarking, and progress monitoring. When possible, proposals should build on and reuse public results from relevant previous funded actions. Communicable results should be shared with the European R&D community through the AI-on-demand platform. The project selected in this topic should link to the resources offered by the AI Factories and the Data Labs. Where relevant, it could also establish links with European companies developing frontier AI models. All proposals are expected to allocate tasks for cohesion activities with the European Partnership on AI, data, and robotics (ADRA) and the CSA HORIZON-CL4-2025-03-HUMAN-18: GenAI4EU central Hub. null Activities are expected to start at TRL 2 and achieve TRL 4 by the end of the project – see General Annex B.

[1] COM(2025)723 Apply AI Strategy

Type of action: HORIZON Research and Innovation Actions

Programme period: 2021 - 2027

Priorities: AI, DigitalAgenda

"> General conditions

1. Admissibility Conditions: Proposal page limit and layout

described in Annex A and Annex E of the Horizon Europe Work Programme General Annexes.

Proposal page limits and layout: described in Part B of the Application Form available in the Submission System.

2. Eligible Countries

described in Annex B of the Work Programme General Annexes. A number of non-EU/non-Associated Countries that are not automatically eligible for funding have made specific provisions for making funding available for their participants in Horizon Europe projects. See the information in the Horizon Europe Programme Guide .

3. Other Eligible Conditions

In order to achieve the expected outcomes, and safeguard the Union’s strategic assets, interests, autonomy, and security, it is important to avoid a situation of technological dependency on a non-EU source, in a global context that requires the EU to take action to build on its strengths, and to carefully assess and address any strategic weaknesses, vulnerabilities and high-risk dependencies which put at risk the attainment of its ambitions. For this reason, participation is limited to legal entities established in Member States, Iceland and Norway and the following additional associated countries: Canada, Israel, Republic of Korea, New Zealand, Switzerland, and the United Kingdom. In addition, entities established in third countries which may become associated to Horizon Europe during 2026 and 2027 may be eligible to parti