Artificial Intelligence in European Project Evaluation: Opportunities and Governance Challenges

Artificial intelligence is already becoming part of the European funding ecosystem. From proposal preparation to project management, AI-powered tools are transforming the way companies, consultants, and public institutions work. The question is no longer whether AI should be used, but how it can be used responsibly, transparently, and in line with the principles governing European research and innovation programmes.

Against this backdrop, the European Association of Innovation Consultants (EAIC) has published its "Position on the Responsible and Governed Use of AI in Proposal Evaluation", outlining a series of recommendations to ensure that the use of artificial intelligence in proposal evaluation preserves trust, fairness, and the overall quality of the assessment process. This paper can be seen here: https://www.eaic.eu/uploads/1/3/1/2/131288413/eaic_position_on_the_responsible_and_governed_use_of_ai_in_proposal_evaluation_.pdf

Evaluation Must Remain a Human Responsibility

The document delivers a clear message: AI can support the evaluation process, but it must never replace the expert judgement of evaluators.

Evaluating R&D and innovation projects involves assessing complex aspects such as scientific excellence, the level of innovation, expected impact, and implementation quality. These decisions require context, experience, and critical judgement—qualities that cannot be fully delegated to an automated system.

For this reason, the EAIC advocates a human-in-the-loop approach, whereby the final decision always remains the responsibility of qualified human evaluators.

Transparency and Traceability

Another key recommendation concerns the need for complete transparency whenever AI is used during the evaluation process.

Organisations employing AI should clearly define:

• which tasks are performed by AI;

• which tasks remain exclusively under human responsibility;

• how AI-generated outputs are supervised;

• and what mechanisms are in place to identify and correct errors or potential bias.

Ensuring traceability is essential to maintaining applicants' confidence and safeguarding the integrity of the evaluation process.

Protecting Confidential Information

European project proposals often contain highly sensitive scientific, technical, and commercial information.

The EAIC therefore stresses that any AI tools used during proposal evaluation must fully comply with European data protection legislation while guaranteeing the confidentiality of applicants' information throughout the entire evaluation process.

Preventing Bias and Ensuring Fairness

Artificial intelligence systems may reproduce or amplify biases present in the data on which they have been trained.

Within a competitive public funding environment, this represents a significant risk that must be addressed through appropriate oversight, auditing, and continuous validation procedures.

According to the EAIC, AI should contribute to improving the consistency and objectivity of evaluations rather than introducing new sources of bias or unequal treatment.

AI Should Improve Efficiency, Not Replace Expert Judgement

Artificial intelligence can add considerable value by supporting administrative and analytical tasks such as document organisation, information retrieval, consistency checks, and workflow management.

However, assessing the scientific, technological, and business quality of research proposals should remain firmly in the hands of independent experts who assume full responsibility for their evaluations. This approach is fully aligned with the principles increasingly adopted by European funding institutions to preserve the credibility and integrity of their evaluation systems.

A Trend That Will Shape the Future of European Funding Programmes

The publication of this position paper reflects the rapid evolution of artificial intelligence within the European funding landscape.

Over the coming years, funding organisations are expected to progressively incorporate AI tools to improve the efficiency and consistency of their evaluation processes. Nevertheless, this transformation will only be successful if accompanied by robust governance frameworks based on transparency, human oversight, accountability, and confidentiality.

For organisations participating in European funding programmes, understanding this evolving landscape will become just as important as understanding the funding opportunities themselves. Confidence in the evaluation process will continue to be one of the fundamental pillars of the European research and innovation ecosystem.

Looking Ahead: AI Governance in the European Research Ecosystem

The publication of the EAIC position paper comes at a pivotal moment for artificial intelligence in Europe. The entry into force of the European Union's Artificial Intelligence Act (AI Act), together with the growing adoption of AI across research and innovation activities, highlights the need for governance models that ensure transparency, human oversight, accountability, and public trust.

Proposal evaluation is no exception. The future is likely to see the emergence of hybrid assessment models in which AI supports evaluators by improving efficiency and consistency, while the final responsibility for funding decisions remains firmly with experienced human experts.