An AI Foundation Toxicology Model and Framework to Support Waiving a Second Species in Drug Safety Studies

This two stage Horizon IHI Research and Innovation Action funds the development of a validated AI Foundation Toxicology Model and a standardized weight of evidence framework to support waiving a second species in chronic and sub chronic repeat dose drug safety studies and to advance NAM based regulatory acceptance. Proposals must be multi stakeholder public private consortia including industry, regulators, academia and SMEs and must demonstrate that private members and contributing partners collectively provide at least 45 percent of the action's eligible costs via in kind or financial contributions. The topic expects to fund one project with an indicative contribution of €9,000,000 and has a short proposal deadline of 8 October 2026 and a full proposal deadline of 21 April 2027 submitted via the EU Funding and Tenders Portal.

What it funds

Scope summary

Develop a validated AI-driven New Approach Methodology (AI Foundation Toxicology Model) plus a transparent weight-of-evidence framework, data infrastructure, tools, training and regulatory engagement to enable waiving second-species chronic and sub-chronic repeat-dose toxicology studies (primarily for small molecules) and promote NAM-based regulatory acceptance.

Who can apply:Legal entities established in Horizon Europe Member States and associated countries (includes Canada, Switzerland, the UK, South Korea where association or specific provisions apply). Proposals must be public-private consortia with relevant industry, academic, regulator and SME participation and meet IHI JU rules on industry contributions IHI Call 13 Topic ¹.

1. 1Consortium: multi-partner public-private; must include industry contributors (IHI private members or approved contributing partners) to meet IHI contribution rules.
2. 2Country eligibility: applicants from EU Member States and Horizon Europe associated countries are eligible for funding; non-associated third-country participation may be possible only in exceptional, justified cases.
3. 3IHI-specific finance rule: full proposals must reflect that at least 45% of total eligible costs and related additional activities are contributed by industry members/contributing partners (IKOP, IKAA, FC).

How much (budget and timing)

Grant size and deadlines

Typical award for this topic:This specific topic HORIZON-JU is budgeted for a single expected grant with a typical IHI JU contribution set at €9,000,000 (min/max contribution listed as €9,000,000).

Key application notes

First-stage short proposals are limited in length and consortia invited to stage two will integrate with the pre-identified industry consortium for full proposals; full proposals must include the IHI-specific Annex to the budget and Type of Participants and meet the 45% industry contribution eligibility rule before evaluation. ¹

Evaluation and format:Two-stage IHI JU RIA procedure: stage 1 short proposal (screening for excellence/impact/implementation), stage 2 invited full proposal (detailed work plan, budget, IHI annexes). Page limits and templates are available in the submission system and IHI guidance ¹.

Eligibility highlights

1. 1Legal entities in EU Member States and Horizon Europe associated countries are eligible to receive funding when meeting standard Horizon Europe rules.
2. 2Industry involvement is mandatory in practice: private members or approved contributing partners must supply a large share of in-kind/financial contributions (45% threshold).
3. 3Non-associated third-country participants may be included only if essential and justified; check country-specific association status before applying.

Where to apply / further reading

Submit via the EU Funding & Tenders Portal topic page and use the IHI application templates; see the topic page for the call text, templates and submission entry Topic page.

Footnotes

1. 1See IHI JU Guide for Applicants and Call documents for templates, annex requirements (including Annex: Budget and Type of Participants) and detailed rules on the 45% industry contribution: IHI Guide for Applicants.

Opportunity and Call Summary

Call:Innovative Health Initiative JU Call 13. Topic reference: HORIZON-JU. Type of action: HORIZON JU Research and Innovation Actions (RIA). Submission procedure: two-stage (short proposal then full proposal). Deadlines: first-stage short proposal deadline 2026-10-08; second-stage full proposal deadline 2027-04-21. This topic seeks to fund one large consortium project (expectedGrants = 1) to develop an AI Foundation Toxicology Model (a NAM) and a standardised weight-of-evidence framework to support regulatory waiving of second species chronic and sub-chronic repeat-dose toxicology studies for small molecule medicines and to enable longer-term expansion to other modalities (e.g., oligonucleotides).

Expected grant size:The topic-specific indicative contribution to a single selected project is €9,000,000 (minContribution = maxContribution = €9,000,000) and the call-level topic budget is shown in the IHI call budget mapping ¹.

What the action will deliver

The project must deliver all of the following outcomes:a validated AI Foundation Toxicology Model producing transparent probabilistic predictions for likelihood of additional safety-relevant information from testing in a second species and for missed toxicity/organ-specific findings/NOAEL divergence; a standardised, reproducible weight-of-evidence decision support framework integrating model outputs and orthogonal evidence; functional tools, templates and training materials addressing explainability, provenance, governance and regulatory alignment; a regulatory strategy and demonstration of regulator engagement and evidence sufficient to support pathway(s) toward regulatory endorsement/validation of second species waivers; and a sustainability and evolution plan for hosting, maintenance, continuous data harmonisation and future expansion beyond small molecules.

Project activities required (representative):consolidation/curation of industry and public toxicology datasets including CDISC SEND and study reports; creation of a common/federated data model and database with privacy preserving approaches and blinding options; AI method selection, development and optimisation (multimodal representation learning, federated learning, probabilistic supervised models, uncertainty quantification, mechanistic modules, scenario simulation, explainable AI and provenance tracking); rigorous model validation, benchmarking and formal testing with pre-defined performance metrics; development and piloting of a standardised weight-of-evidence framework and decision support tools; use-case driven regulator engagement and validation activities; training materials, templates and governance documents; and a sustainability plan that addresses continuous learning and modular updates.

Context and rationale:ICH M3(R2) historically requires two-species toxicity testing for small molecules. Scientific advances (NAMs, in vitro systems, multi-omics and AI) and evolving policy (EU roadmap to phase out some animal testing, EMA NAM initiatives, FDA ISTAND roadmap, NC3Rs projects) create an opportunity to evidence NAM-enabled waivers. The topic targets reduction of animal use, faster development timelines, lower costs and enhanced regulatory consistency globally.

Detailed Eligibility, Consortium and Funding Rules

This is an IHI JU two-stage RIA topic. Short proposals (Stage 1) are limited to 20 pages; invited full proposals (Stage 2) to 50 pages for Part B. The submission system will generate Part A. Proposals must follow the IHI templates and upload required IHI annexes. The topic text and IHI Work Programme specify IHI-specific conditions including the minimum industry contribution threshold (see below). The call follows Horizon Europe legal and financial rules and the IHI JU Work Programme and Guide for Applicants.

IHI JU industry contribution requirement:To be eligible, a full proposal must demonstrate that at least 45% of the action’s total eligible costs (plus costs for action-related additional activities) will be provided collectively by IHI private members (their constituent or affiliated entities) and/or IHI contributing partners, using a combination of in-kind contributions to operational activities (IKOP), in-kind contributions to additional activities (IKAA) and/or financial contributions (FC). The Annex to the Budget and Type of Participants must be completed to demonstrate this threshold at full proposal stage; IHI strongly recommends aiming at 50% industry contribution as a prudent margin.

Where participants can be located:Legal entities established in EU Member States and Horizon Europe associated countries are eligible to receive funding in this call. The topic text explicitly notes that legal entities based in all countries associated to Horizon Europe, including Canada, Switzerland, the UK and South Korea, are eligible to receive funding under this call ¹. Exceptional funding for essential third-country participants may be requested and justified; this is considered case-by-case per Horizon Europe rules.

Administrative and submission requirements

Proposals must be prepared and submitted electronically via the Funding & Tenders Portal. Use the IHI RIA Part B templates (stage 1 short proposal and stage 2 full proposal) and the IHI-specific annexes available in the submission system. The Annex: Budget and Type of Participants is compulsory at full proposal stage. Other compulsory annexes for full proposals include the Declaration of In-Kind Contribution Commitment and the IKAA annex where applicable; an Essential Information for Clinical Studies annex is required if the proposal includes clinical studies (otherwise a statement that the proposal includes no clinical studies must be uploaded). The IHI Guide for Applicants provides full procedural detail and templates.

1. 1Files and templates: Stage 1 – Part A (auto) and Part B short proposal template (20 pages); Stage 2 – Part A (auto) and Part B full proposal template (50 pages) and IHI annexes.
2. 2Compulsory annexes at full proposal stage: Annex: Budget and Type of Participants; Declaration of In-Kind Contribution Commitment; Annex: IKAA when IKAA included; Annex: Essential information for clinical studies (or negative statement).
3. 3Submission system: Funding & Tenders Portal (electronic submission).
4. 4Formatting rules: follow Part B template fonts, margins and page limits; excess pages will be ignored by evaluators.
5. 5Two-stage schedule: after Stage 1 evaluation the highest-ranked stage 1 applicant under this topic is invited to merge with pre-identified industry consortium and submit a Stage 2 full proposal.

Evaluation and selection

Evaluation follows IHI JU and Horizon Europe rules. Independent experts assess proposals against the three award criteria (Excellence, Impact, Quality and efficiency of implementation) with the standard thresholds (normally ≥3 for each criterion and total ≥10/15 unless otherwise specified). Stage 1 short proposals are evaluated to select a single applicant consortium to proceed to Stage 2 and merge with the pre-identified industry consortium; Stage 2 full proposals are evaluated and ranked; funding decisions are taken from ranked list constrained by topic budget. IHI RIA evaluation form and guidance are provided with call documents; applicants are strongly advised to use the IHI evaluation form to self-assess before submission.

Project duration and reporting:Project duration should be proportionate to the work proposed; realistic timelines to achieve model training, validation, regulator piloting and sustainability planning are expected (multi-year programmes are likely). Funded projects will follow Horizon Europe / IHI reporting, deliverable and audit rules under the Model Grant Agreement.

Key technical expectations and science requirements

Applicants must demonstrate technical readiness across data, AI, mechanistic toxicology and regulatory science. Key technical elements expected include: dataset identification and curation (CDISC SEND and unstructured reports), common/federated data model design, privacy preserving techniques and blinding strategies for structural data, multimodal representation learning (including contrastive methods), feature extraction pipelines for each data modality (study reports, chemical structure, omics, mechanistic assays), federated learning approaches, probabilistic predictive modelling (Bayesian neural networks, Gaussian processes, calibrated ensembles), uncertainty quantification, mechanistic-informed modules (knowledge graphs, causal inference), scenario simulation engines (Monte Carlo, generative counterfactuals), explainable AI and provenance tracking, formal benchmarking against held-out datasets and robust performance metrics (accuracy, calibration, robustness, explainability), and multi-stakeholder validation with regulators and industry.

1. 1Data sources and formats: structured SEND datasets, unstructured study reports, chemistry, PK/exposure, omics, in vitro NAMs, prior consortia outputs, public regulatory databases; data harmonisation and FAIR principles required.
2. 2Privacy and IP: federated analyses, privacy-preserving techniques, clear IP and access rules; consider anonymisation, blinding, structural embeddings with noise, and contractual arrangements for contributing partners/private members.
3. 3Model outputs: probabilistic predictions (0–100) for the likelihood that second species or longer-term studies will detect novel findings, organ-specific effects, or NOAEL divergence; calibrated and explained outputs for regulator review.
4. 4Weight-of-evidence framework: structured reproducible decision support integrating model outputs, orthogonal NAMs and literature to assess evidence quality, relevance, consistency and uncertainty.
5. 5Regulatory engagement: planned regulator interaction, predefined use-cases and performance targets, alignment with EMA/FDA/ICH expectations (e.g., ICH M3(R2) guidance), and potential ISTAND or analogous submissions.

Risk management, ethics and animal welfare

The topic explicitly aims to reduce animal testing and contribute to 3Rs implementation. Proposals must address data provenance, data quality, ethics and patient/public engagement. The consortium must include an ethics plan and describe how human data, if used, will meet GDPR and local consent requirements. Animal welfare and justification for any new in vivo studies must be explicit and follow Directive 2010/63/EU and applicable 3R principles.

Templates and application forms

The IHI JU Part A is generated by the Funding & Tenders Portal. Part B templates (short and full) and all IHI annex templates referenced below are available to download in the submission system. Applicants must use the IHI ‘Part B short proposal’ template for stage 1 (20 pages) and the ‘Part B full proposal’ template for stage 2 (50 pages). Required annex templates: Annex: Budget & Type of Participants; Annex: Declaration of In-Kind Contribution Commitment; Annex: IKAA (if IKAA included); Annex: Contributing Partner application letter(s) (if contributing partners included in a single-stage call); Annex: Essential Information for Clinical Studies (if clinical studies included) and any ethics annexes if needed.

1. 1Download Part B short proposal template (IHI RIA/SP) and Part B full proposal template (IHI RIA/FP) from the Submission System.
2. 2Complete Annex to the Budget & Type of Participants at full proposal stage to demonstrate the 45% industry contribution.
3. 3Complete and sign IHI Declaration of In-Kind Contribution Commitment for the consortium where applicable.
4. 4If including contributing partners in a single-stage proposal, upload signed contributing partner application letter(s) as part of the proposal (CP Annex).
5. 5If including IKAA, upload the IKAA annex describing the additional activities and their value.

Category extraction and structured answers

Eligible Applicant Types

The topic welcomes multi-stakeholder consortia. Eligible applicant types include industry (large companies, mid-caps and SMEs), pharmaceutical companies, biopharmaceutical companies, medical technology companies, research organisations and universities, research and technology organisations (RTOs), public authorities, regulatory bodies, notified bodies, healthcare providers, health technology assessment bodies, patient and citizen organisations, NGOs, charities and foundations, and other legal entities eligible under Horizon Europe. For IHI JU specifically, IHI private members (COCIR, EFPIA including VaccinesEurope, EuropaBio, MedTech Europe) and their constituent/affiliated entities, and IHI contributing partners are core participants for industry contribution and eligibility calculations.

Funding Type

Primary financial mechanism:grant (Horizon JU Action Grant, budget-based). The IHI JU grant follows the Horizon Europe Model Grant Agreement (HE MGA) and IHI JU specific rules.

Consortium Requirement

Consortium is required. This is a two-stage call where stage 1 short proposals are submitted by applicant consortia; the successful stage 1 applicant consortium will be invited to submit a stage 2 full proposal merged with the pre-identified industry consortium. The full consortium at stage 2 should include academic, SME, industry, regulator and other relevant partners. A multi-partner public-private consortium is expected.

Beneficiary Scope (Geographic Eligibility)

Primary geographic eligibility:entities established in EU Member States and countries associated to Horizon Europe are eligible to receive funding. The topic text explicitly lists Canada, Switzerland, the UK and South Korea among countries associated and eligible to receive funding under this call. Third-country participants not automatically eligible may exceptionally be funded if essential, per Horizon rules.

Target Sector

Target sectors and themes:health (pharmaceuticals, toxicology, regulatory science), artificial intelligence and machine learning, new approach methodologies (NAMs), in vitro systems, multi-omics, pharmacokinetics, safety pharmacology, data science, digital health, medical devices, biotech, clinical development, and animal welfare/3Rs.

Mentioned Countries

Explicitly mentioned countries in the topic and metadata:Canada, Switzerland, United Kingdom, South Korea. Region: EU and Horizon Europe associated countries generally.

Project Stage

Expected maturity:development, validation and demonstration. Activities must include foundation model development (pre-training and fine-tuning), rigorous validation and benchmarking, regulatory validation use-cases, and sustainability planning. TRL expected to progress from research/TRL 4-6 to validation/demonstration/TRL 7+ by project end.

Funding Amount

Topic indicative contribution:€9,000,000 for one expected grant under this topic (minContribution = maxContribution = €9,000,000). Call-level budget shown in the IHI topic mapping. Applicants should request amounts aligned to the scope and be aware of the IHI-specific in-kind contribution requirements.

Application Type

This is an open two-stage call. Stage 1:open short proposal submission. Stage 2: invited full proposal by the successful Stage 1 applicant merged with the pre-identified industry consortium. Submission via the Funding & Tenders Portal.

Nature of Support

Financial support (grant funding) plus expectations for non-monetary ecosystem contributions from industry (in-kind operational activities IKOP, in-kind additional activities IKAA, and possible financial contributions FC paid/received).

Application Stages

Two stages:Stage 1 short proposal (20 pages) evaluation; top-ranked stage 1 applicant invited to Stage 2; Stage 2 full proposal (50 pages) evaluation leading to ranking and funding decision.

Success Rates

ExpectedGrants for this topic:1. The call is highly competitive; historical success rates for IHI large RIA topics are typically low (single-digit to low double-digit percent) depending on the number and quality of submissions and available budget. Exact success probability cannot be pre-specified and depends on the number of high-quality applications in the call.

Co-funding Requirement

Yes. The IHI JU-specific eligibility rule requires that industry (IHI private members and/or contributing partners) provide at least 45% of the action’s eligible costs and costs for action-related additional activities via IKOP, IKAA and/or FC paid; applicants are advised to aim for 50% as prudent margin. The full proposal must include the Annex to the Budget and Type of Participants documenting this contribution.

Templates and required application documents (practical)

Key templates and annexes available in the Submission System and on the IHI website. Applicants must follow the precise Part B structure and page limits. The following IHI JU templates/annexes are required or may be required depending on content:

1. 1Proposal Part A (administrative) – generated in the Funding & Tenders Portal.
2. 2Proposal Part B short proposal template (Stage 1) – 20 pages limit, required for two-stage procedure.
3. 3Proposal Part B full proposal template (Stage 2) – 50 pages limit, required when invited to full proposal.
4. 4Annex: Budget and Type of Participants – compulsory for full proposals; required to demonstrate 45% industry contribution.
5. 5Annex: Declaration of In-Kind Contribution Commitment – compulsory for full proposals.
6. 6Annex: In-Kind Contributions to Additional Activities (IKAA) – compulsory if IKAA are included (full proposals).
7. 7Annex: Contributing Partner application letter(s) (CP Annex) – compulsory in single-stage calls if contributing partners participate; upload as part of submission.
8. 8Annex: Essential Information for Clinical Studies – compulsory if clinical studies are included; otherwise upload a declaration that no clinical studies are included.
9. 9Ethics annex – optional but required if complex/serious ethics issues cannot be described within Part A character limits.

Mandatory IHI financial annexes include the IHI-specific Excel budget annexes that capture IKOP, IKAA, non-EU IKOP, FC paid and received, and checks for 45% threshold and non-EU IKOP limits. The IHI finance team can be contacted for annex completion questions.

How to prepare a strong proposal — practical tips

Address the topic scope and expected outcomes precisely:define representative use cases, the data and datasets you will secure, the model architecture, explainability strategy, uncertainty quantification, validation plan and regulatory engagement plan. Demonstrate consortium readiness: data access agreements, legal/IP plans, experience in toxicology and AI, evidence of privacy-preserving/federated approaches and compute resources, and a concrete sustainability/business model for long-term maintenance. Define clear performance metrics and decision thresholds that regulators and industry co-develop. Provide evidence of early regulator engagement and propose timelines for regulator validation activities.

1. 1Data: show concrete lines of commitment (letters, MOUs) for SEND datasets and study reports; describe how unstructured reports will be harmonised.
2. 2Methodology: describe the pre-training, fine-tuning, multimodal architecture, probabilistic calibration and uncertainty measures.
3. 3Validation: define held-out benchmarking datasets, performance metrics (ROC, calibration, false-negative risk), stress tests and counterfactual scenario evaluations.
4. 4Regulatory: include a regulator interaction plan, mention potential ISTAND or EMA interactions, and describe how outputs map to regulatory decision-making.
5. 5Governance and sustainability: propose hosting, access policies (public vs controlled), model update and continuous learning workflows, and cost/revenue sustainability approaches.

Conclusion — What is this opportunity about and how to explain it

This IHI JU two-stage RIA topic funds a large public-private consortium to create and validate an AI Foundation Toxicology Model and associated standardised weight-of-evidence framework that can predict when second species chronic or sub-chronic testing for small molecules is unlikely to yield additional safety-relevant information. The deliverables are a rigorously validated, explainable probabilistic model trained on curated multi-source toxicology and NAM datasets; a decision-support weight-of-evidence framework and tools for industry and regulators; regulator engagement and evidence to support pathway(s) for acceptance of second species waivers; and a sustainability plan for the foundation model and ecosystem.

This is a high-value, large-scale RIA expected to require multi-disciplinary expertise across pharma, toxicology, AI, data engineering, regulatory science and governance. Proposers must plan early and secure legal agreements for sensitive data, define robust AI methodologies aligned with regulatory expectations, and provide a clear regulatory evidence generation and engagement plan. The IHI-specific requirement for substantial industry contribution (45% minimum) is mandatory for eligibility and reflects the public-private partnership model.

Quick reference checklist

• Confirm two-stage submission procedure and prepare to submit a concise stage 1 (20 pages) Part B short proposal by 2026-10-08.
• Ensure consortium includes partners able to provide the IHI JU industry contribution (IKOP/IKAA/FC) to reach ≥45% (aim for 50%).
• Secure letters/MOUs from data providers and IHI private members/contributing partners where possible for stage 1; full supporting documents will be needed for stage 2.
• Plan early regulator engagement and define targeted use-cases and performance criteria.
• Prepare Part B to follow IHI templates exactly and have Part A administrative fields completed in the Portal.
• At full proposal stage, complete Annex: Budget & Type of Participants and any relevant IHI annexes (IKAA, Declaration of in-kind contribution, CP Annex if applicable).

Footnotes

1. 1IHI JU topic listing and call documents are available on the Funding & Tenders Portal and IHI website; see topic page for HORIZON-JU and the IHI Call 13 documentation IHI Call 13 topics and documents per the call text and budget mapping.

What This Opportunity Is About

This call funds the development of a validated Artificial Intelligence (AI) Foundation Toxicology Model and a standardized weight-of-evidence framework to enable waiving the requirement for a second species in chronic and sub-chronic drug safety studies for small molecules and other modalities like oligonucleotides ¹. The project aims to replace, reduce, and refine animal testing (3Rs principles) by using AI-driven New Approach Methodologies (NAMs) to predict whether a second species study will provide additional safety-relevant information, such as missed toxicity or divergent NOAEL values ².

Key Expected Impacts

1. 1Faster and more informed decision-making through AI-driven NAMs and rapid data processing
2. 2Regulatory adoption of a NAM-enabled second species waiver model and weight-of-evidence framework
3. 3Significant reduction in animal use, accelerated timelines, and lower costs for the European health industry
4. 4Improved public health by delivering safe and effective medicines faster using validated NAMs
5. 5Contribution to EU Directive 2010/63/EU on animal protection and global consistency in regulatory decision-making

Who Should Apply

The ideal consortium must include pharmaceutical companies, regulatory agencies (e.g., EMA, FDA), academic institutions, SMEs, and patient advocacy groups. A critical eligibility requirement is that private members (e.g., pharmaceutical companies) and/or contributing partners must collectively provide at least 45% of the action's total eligible costs through in-kind contributions (IKOP), financial contributions (FC), or in-kind contributions to additional activities (IKAA) ³.

Consortium Composition Requirements:Proposals must demonstrate a multi-stakeholder approach including industry (pharma/biotech), regulators, academia, and SMEs to ensure broad ecosystem adoption and regulatory confidence ⁴.

Eligibility Highlights

Project Scope & Deliverables

The project must deliver five core outcomes:(1) A validated AI Foundation Toxicology Model providing probabilistic predictions for second species waivers; (2) A standardized weight-of-evidence framework for reproducible assessment; (3) Functional tools, templates, and training materials for real-world implementation; (4) Enhanced stakeholder confidence through regulatory validation; and (5) A sustainability plan for long-term model evolution and ecosystem adoption ⁵.

Core Technical Tasks:1. Consolidate and evaluate diverse data sources (CDISC SEND, unstructured reports, omics, PK data); 2. Select and optimize AI methods (federated learning, Bayesian neural networks, multimodal representation); 3. Develop and validate the model across study durations and species; 4. Establish the weight-of-evidence framework; 5. Prepare for regulatory uptake (e.g., ICH M3(R2) revision, EMA/FDA endorsement) ⁶.

Funding & Budget Details

Deadlines & Submission Procedure

This is a **two-stage call**. The first stage deadline for short proposals is **October 8, 2026**, and the second stage deadline for full proposals is **April 21, 2027** ⁷. Short proposals (Stage 1) have a page limit of 20 pages, while full proposals (Stage 2) have a limit of 50 pages.

Submission Timeline:Stage 1: Submit short proposal by Oct 8, 2026. If selected, proceed to Stage 2. Stage 2: Submit full proposal by April 21, 2027. Submission is via the EU Funding & Tenders Portal ⁸.

Regulatory Strategy & Long-Term Goals

Applicants must develop a regulatory strategy to gain endorsement from authorities like EMA and FDA (potentially via the ISTAND program) by the project's end. The longer-term goal, beyond the action's scope, is to support the revision of the **ICH M3(R2)** regulatory guideline to incorporate these AI-driven outcomes. The project should also plan to broaden the model's scope to enable single-species chronic testing waivers for other modalities like oligonucleotides ⁹.

Key Regulatory References

1. 1ICH M3(R2): Nonclinical safety studies for human clinical trials
2. 2ICH S6: Preclinical safety evaluation of biotechnology-derived pharmaceuticals
3. 3EU Directive 2010/63/EU: Protection of animals used for scientific purposes
4. 4EU AI Act (Regulation 2024/1689): Governance of high-risk AI systems
5. 5ALTAI: Assessment List for Trustworthy Artificial Intelligence

Footnotes

1. 1Official Call Text: HORIZON-JU, available at EU Funding Portal.
2. 2Expected Impact & Outcome details from the call documentation, emphasizing NAM-enabled waivers and 3Rs principles.
3. 3IHI JU Annex on Declaration of In-Kind Contribution Commitment, requiring 45% industry contribution (IKOP + FC paid + IKAA) IHI Declaration.
4. 4IHI Stakeholders Overview and Type of Participants Annex, detailing required consortium composition IHI Annex.
5. 5Expected Outcomes section of the call text, listing the 5 core deliverables including the AI model and weight-of-evidence framework.
6. 6Scope section detailing the 5 proposal tasks: data consolidation, AI method selection, model validation, framework establishment, and regulatory uptake.
7. 7Deadline Dates listed in the call actions: 2026-10-08 (Stage 1) and 2027-04-21 (Stage 2).
8. 8Submission procedure details from the Funding & Tenders Portal Online Manual and IHI JU Guide for Applicants.
9. 9Regulatory Strategy section mentioning ICH M3(R2) revision and ISTAND program engagement with EMA/FDA.