KETs2026 reloaded

24 Sept 2026 | Germany

Project cooperationUpdated on 11 August 2026

Seeking an eligible European consortium for a healthcare AI challenge and evaluation workstream

Internal Medicine Specialist | Clinical AI Safety and Evaluation Researcher at Göktuğ Özkan — Independent Clinical-AI Research

Istanbul, Türkiye

About

Target topic

HORIZON-CL4-2027-04-DIGITAL-EMERGING-04 — Apply AI: Challenge-Driven AI Innovation Booster in Apply AI prioritised sectors

Proposed cooperation

The target topic calls for a challenge-driven project in one strategic sector, with healthcare guaranteed portfolio coverage if a proposal passes all thresholds. It requires a multi-stage competition, user-industry core partners, SME participation, measurable KPIs, benchmarking, and a solid evaluation methodology.

SAFE-CLINAI is proposed as the clinical challenge-design and evaluation layer for such a consortium. It would support:

• synthetic, traceable clinical challenge cases and controlled adversarial variants;

• clinician-authored evaluation criteria and blinded review workflows;

• metrics for clinical correctness, harmful omission, unsupported confidence, evidence traceability, robustness, and workflow usefulness;

• repeatable regression testing across challenge stages;

• transparent scoring, adjudication, and evaluation governance;

• translation of evaluation findings into actionable feedback for participating AI teams.

What I bring

I am an Internal Medicine Specialist and clinical-AI safety researcher working in an individual capacity. TurKMedBench provides a research seed for the proposed method: a historical hematology study used 24 synthetic cases across five models, producing 120 responses for structured failure analysis. It is not a clinically validated product, and the multi-specialty expansion remains controlled work in progress.

I can contribute to clinical use-case definition, challenge specifications, synthetic case design, safety and uncertainty criteria, evidence-traceability scoring, clinician-review workflows, and evaluation governance.

Partners sought

• An eligible European coordinator with Horizon Europe RIA and FSTP experience

• Healthcare user-industry organisations able to act as core partners and support future uptake

• AI SMEs, model developers, data/platform providers, and challenge infrastructure teams

• Hospitals, universities, and clinical research organisations from eligible countries

• Experts in ethics, regulatory strategy, data governance, exploitation, communication, and impact

Collaboration and eligibility note

The concept is at consortium-building stage; no institutional or partner commitment is implied. The current topic text does not list Türkiye among eligible beneficiary countries. Any role for me or a Türkiye-based organisation must therefore be structured and confirmed as legally compliant under the final topic conditions and applicable employment rules before commitment.

Contact

Göktuğ Özkan, MD

Internal Medicine Specialist | Clinical AI Safety and Evaluation Researcher

drozkan2@gmail.com

Stage

  • Early

Topic

  • AI-GenAI /Data/Robotics

Type

  • Partner seeks Consortium

Similar opportunities