Advancing trustworthy AI in CCAM – Key policy recommendations from the AIthena project

As Europe accelerates towards a future of connected, cooperative and automated mobility (CCAM), it becomes a critical policy imperative to ensure that AI-driven systems are trustworthy, transparent, safe and human-centred. AIthena’s deliverable D6.3, ‘Lessons learned and policy recommendations’, offers a comprehensive synthesis of insights generated throughout the project, consolidating technical findings, ethical analyses and real-world testing outcomes into a forward-looking policy framework.

Drawing on contributions from domain experts across research organisations, industry partners and testing facilities, D6.3 provides evidence-based policy guidance tailored for European institutions, national authorities, standardisation bodies and stakeholders in the CCAM ecosystem.

The deliverable is aligned with AIthena’s mission to strengthen trustworthy AI for CCAM, integrating explainability (XAI), robust data governance, ethical transparency and safety-driven design into future regulatory and operational frameworks.

Key policy recommendations

The main policy recommendations put forward in D6.3 are summarised below, structured around the most pressing governance needs for AI-based CCAM systems.

1. Strengthen regulatory alignment for trustworthy AI.

D6.3 emphasises the need for harmonised regulatory frameworks that can keep pace with the rapid evolution of AI technologies in the transport sector. This includes:

  • Aligning CCAM-specific AI regulation with broader EU-level initiatives (e.g. the AI Act and the Data Governance Act).
  • Ensuring consistent requirements for transparency, accountability and explainability across national contexts.
  • Promoting a shared European understanding of the risk levels associated with automated mobility systems.

Such harmonisation is essential to prevent fragmented implementation and ensure that automated mobility technologies can be deployed safely across borders.

2. Enhance data governance and access mechanisms.

Robust data governance frameworks are essential for trustworthy and effective CCAM. D6.3 recommends the following:

  • Establishing clear rules for data collection, sharing and usage, particularly in safety-critical scenarios.
  • Encouraging the creation of secure data spaces that respect privacy and facilitate cross-sector collaboration.
  • Supporting standardised data formats to improve interoperability among vehicles, infrastructure and testing environments.

These measures are designed to enable innovation while safeguarding ethical and legal principles.

3. Promote explainable and transparent AI (XAI).

Due to the high-stakes nature of automated mobility systems, D6.3 emphasises the importance of explainability as an essential regulatory requirement. Recommendations include:

  • Mandating interpretable AI models for safety-related decisions, including real-time explanations where feasible.
  • Supporting research into XAI methods tailored to the complexity of CCAM.
  • Requiring the documentation and communication of AI decision processes for stakeholders, auditors and end users.

This emphasis on XAI reflects AIthena’s broader goal of building public trust in AI-driven mobility systems.

4. Support comprehensive testing, validation and certification frameworks.

AIthena’s testing partners provided valuable input on the challenges of validating AI behaviours in different conditions. D6.3 calls for:

  • Developing harmonised testing protocols and certification schemes for AI-based CCAM systems.
  • Integrating virtual and real-world testing environments to capture edge cases.
  • Strengthening cross-country collaboration in testbed infrastructure and knowledge exchange.

A robust and coordinated testing landscape will help ensure consistent safety and performance standards across Europe.

5. Integrate ethical and human-centric design principles.

D6.3 emphasises that AI systems in the field of mobility must reflect European societal values. Policy actions include:

  • Embedding ethics assessments into the entire CCAM development lifecycle.
  • Ensuring that AI outputs respect fairness, non-discrimination and user autonomy.
  • Encouraging stakeholder engagement to align technological development with public expectations.

The recommendations foreground human-centricity in order to preserve societal trust and reinforce responsible innovation.

6. Foster skills development and capacity building.

To sustain long-term progress, D6.3 recommends that policymakers:

  • Support training initiatives for engineers, data scientists, regulators and mobility operators.
  • Develop interdisciplinary curricula combining AI expertise with knowledge of ethics, law and mobility systems.
  • Promote international knowledge exchange networks.

This investment in human capital is essential to keep Europe at the forefront of innovation in trustworthy AI mobility.

7. Encourage collaboration throughout the CCAM ecosystem.

AIthena’s findings demonstrate the value of close collaboration between sectors. D6.3 therefore recommends the following:

  • Facilitating collaboration between industry, academia, certification bodies and public authorities.
  • Supporting cross-project integration and alignment with parallel Horizon Europe initiatives.
  • Building long-term ecosystems for knowledge transfer and innovation.

Collaboration is seen as essential for scaling up safe and interoperable CCAM systems.

You will find the full AITHENA D6.3 report at D6.3 Lessons learned and policy recommendations