CLARA harnesses artificial intelligence, machine learning, and hybrid supercomputing to advance biotech, biomedical, and clinical research. As such endeavors often depend on sensitive or proprietary data and powerful emerging technologies, ethical and regulatory considerations are integral to its activities.
CLARA conducts all scientific work with methodological rigor, transparency, and accountability, upheld through the following commitments:
CLARA deploys AI responsibly and in line with applicable regulations, including the EU AI Act. Systems are risk-classified under the Act, and high-risk applications undergo conformity assessment supported by technical documentation, data governance records, and a logged risk-management process. Across the lifecycle, controls evaluate models for accuracy, robustness, and bias, document training data and known limitations, and maintain human oversight and logging so that outputs remain traceable and contestable. AI for healthcare or medical use is additionally assessed against the Medical Device Regulation (MDR), and deployed systems are monitored and periodically re-evaluated for performance drift, emerging risks, and regulatory changes.
CLARA has adopted a policy establishing AI-compliance standards and mandates all CLARA members to ensure compliance through training and cross-compliance. EU AI Act requirements are being adopted progressively, in line with the Act's phased application and CLARA's implementation progress.
CLARA protects personal data in full compliance with the General Data Protection Regulation (GDPR). Processing rests on a documented lawful basis, and biological and clinical data are handled under purpose limitation, data minimization, and storage limitation. Data are pseudonymized or anonymized where feasible, with privacy-by-design and privacy-by-default applied and Data Protection Impact Assessments (DPIAs) conducted for high-risk processing. Data are safeguarded through technical and organizational measures including encryption in transit and at rest, role-based access controls, audit logging, and secure processing environments. Data-subject rights are upheld, transfers and third-party processing are governed by appropriate safeguards, and privacy-preserving techniques such as federated learning and differential privacy are employed where applicable.
CLARA has established data-protection standards in full conformity with the GDPR, along with privacy policy, cookie policy.
CLARA identifies, discloses, and manages actual, potential, and perceived conflicts of interest so that research decisions remain objective; researchers and staff must declare financial interests, funding sources, commercial relationships, advisory roles, and personal connections that could affect their judgment. Declared conflicts are managed through procedures such as recusal, independent oversight, or withdrawal, and funding and sponsorship are disclosed transparently so that results are interpreted on the evidence alone.
CLARA has adopted conflict-of-interest standards so that research decisions remain objective.
CLARA designs AI to serve people, prioritizing human oversight, fairness, transparency, and the well-being of patients and communities, so that technology supports rather than replaces human judgment and clinicians retain final responsibility for decisions. Systems are explainable and interpretable, bias is tested for and mitigated to ensure equitable outcomes, and patients, clinicians, and the wider community are engaged in shaping how the technology is developed and used.
CLARA adheres to the “Do No Significant Harm” (DNSH) principles, ensuring its research and operations avoid significant harm to environmental and societal objectives. In line with the EU taxonomy, impacts are considered across its six objectives: climate change mitigation and adaptation, sustainable use of water and marine resources, the circular economy, pollution prevention, and the protection of biodiversity and ecosystems.
As large-scale AI and high-performance computing are energy-intensive, CLARA promotes energy-efficient computing and responsible use of supercomputing resources, and assesses the wider societal effects of its work in line with responsible research and innovation.
CLARA ensures the highest standards of professional conduct through adopted policies that translate its ethical principles into day-to-day practice. The Code of Conduct applies to all researchers, staff, partners, and collaborators working within the CLARA ecosystem.
All members are ensured to act with honesty and integrity, comply with applicable laws and policies, safeguard confidential and personal data, respect intellectual property, treat others with dignity and without discrimination, disclose conflicts of interest, and raise concerns about misconduct through the appropriate channels.
These standards are upheld through a structured ethics and compliance program that embeds CLARA's principles into research and operations, combining clear policies, practical controls, and continuous oversight to ensure that ethical and regulatory requirements are met throughout the AI and research lifecycle.
It rests on documented policies and standards, systematic risk assessment (including AI Act conformity), controls and documentation for traceability, ongoing training, and periodic monitoring and review with corrective action where gaps arise.
CLARA maintains clear accountability for ethics and compliance. Oversight rests with CLARA's leadership and a designated Ethics and Compliance Officer, supported by relevant committees and the partner institutions within the CLARA ecosystem.
Governance arrangements include:
This framework ensures that ethical and regulatory commitments are owned at every level and reviewed on an ongoing basis.
For more information see CLARA deliverables (D2.1, D2.5).
Contact: Lalit Patil