Trustworthy AI Framework

Ethical Governance

"Technology is progress only when governed by human values. All AI processes within IMU are secured by universal ethical principles."

AI Ethics:
A Values Revolution

Generative Artificial Intelligence (AI) technologies are radically changing research and education paradigms in higher education, bringing unprecedented ethical responsibilities alongside new opportunities.

This declaration is based on the "Ethical Guide on the Use of Generative AI in Scientific Research and Publications" (May 2024) published by the Council of Higher Education (CoHE/YÖK), as well as the universal ethical frameworks established by the European Commission, UNESCO, and OECD. Our goal is to transform algorithms from "black boxes" into transparent, accountable, and fair systems serving humanity.

3 Pillars of Trustworthy AI

01
Legal Compliance

Full compliance with KVKK (GDPR), AI Act, and relevant regulations.

02
Ethical Principles

Loyalty to fundamental human rights, academic integrity, and CoHE ethics.

03
Technical Robustness

Use of systems that are resilient to errors, secure, and auditable.

Source: IMU Ethical Governance Framework

6 Fundamental Ethical Principles

We apply the following 6 Fundamental Principles in our academic work under the "Duty of Care":

01. Human Autonomy

We use AI systems not as autonomous authors, but as "drafters" supporting human intelligence. The final say in text approval lies with the human.

02. Transparency

"Unexplainable technology is unverifiable technology." AI tools used in research, their purpose, and scope must be declared in the "Methodology" section.

03. Data Governance

Sensitive data or unpublished findings are not uploaded to public AI models (e.g., ChatGPT) without anonymization or data security guarantees.

04. Diversity

We are aware of algorithmic bias. Outputs carrying risks of discrimination based on gender, race, or belief are passed through a critical filter.

05. Scientific Accuracy

Against the risk of AI "hallucinations," all generated data is verified against original sources. Responsibility lies with the user.

06. Authorship Limits

Artificial intelligence cannot be the "Author" of a scientific study; it can only be used as a "Drafter."

Academic Integrity:
Drafting & Curation

The use of AI in academic writing and research brings the risk of "Autonomous Ghost-writing." As the IMU AI Office, we adopt the "Structural Support" approach.

Researchers and students may use AI for language optimization, translation, literature review, and ideation. However, draft texts generated by AI are merely "raw material." Using these outputs as final text without comprehensive editing and verification constitutes an ethical violation.

workspace_premium

IMU Seal of Ethics

"Every piece of knowledge we produce is measured not by the speed of AI, but by human conscience and academic integrity."

The Future is Built with Trust.