Data Protection and AI Transparency Guarantee:
AI-Driven Journal of Interdisciplinary Social Research (AI-JISR) is fully committed to upholding author, reviewer, and reader privacy under the Turkish Personal Data Protection Law No. 6698 (“KVKK”) and the General Data Protection Regulation (“GDPR”). Our journal utilizes artificial intelligence agents operating via NVIDIA API infrastructure during editorial workflows. In strict compliance with double-blind peer review, identifying personal data is never transmitted to AI models; manuscript texts are never used to train external models; and all final binding publication decisions are made exclusively by the human Editor-in-Chief.

1. Data Controller

The data controller is the Editorial Office of AI-Driven Journal of Interdisciplinary Social Research (AI-JISR) (Contact: [email protected]; Web: https://aijisr.org; ISSN: 3062-0988).

2. Categories of Personal Data Processed

  • Identity Information: First name, last name, academic title and degree.
  • Contact Information: Institutional and personal email addresses, institutional affiliation address, country.
  • Academic & Professional Information: Affiliated university/institution, faculty/department, ORCID iD, research areas and disciplinary focus.
  • Manuscript & Workflow Information: Submitted manuscript files, revision documents, contributor statements, ethics approvals, copyright transfer statements, editorial logs.
  • Security & Technical Data: IP address, user account identifiers, cryptographic password hashes (bcrypt), session tokens, system audit trail logs.

3. Provisions on Artificial Intelligence (AI) and NVIDIA API Integration

AI-JISR implements an auditable Multi-Agent AI architecture to streamline manuscript desk checks, field editor classification, and isolated peer review:

A. Anonymization & Double-Blind Integrity

Before manuscript files are parsed for AI analysis, all author names, affiliations, email addresses, and identifying metadata are stripped and masked. Prompts submitted to the NVIDIA API contain solely the scientific manuscript content (title, abstract, methodology, findings, and bibliography). Personal identity information is never shared with AI models under any circumstance.

B. NVIDIA API Infrastructure & Zero Data Retention

AI inference is executed through enterprise NVIDIA NIM / NVIDIA AI Foundation Endpoints (integrate.api.nvidia.com) under enterprise API terms:

  • Submitted academic content is never utilized to train or fine-tune NVIDIA or third-party AI models.
  • Inference is conducted ephemerally; data is not persistently cached or retained on inference microservice servers.
  • All transmission is encrypted end-to-end via TLS 1.3.

C. Human-in-the-Loop Principle

Under GDPR Article 22 and KVKK Article 11, the right not to be subject to a decision based solely on automated processing is fully upheld. AI agents produce solely non-binding advisory evaluations. All definitive publication, revision, and rejection decisions are issued solely by the human Editor-in-Chief.

D. Immutable Audit Trail

Every AI evaluation step, including model identifier (e.g., meta/llama-3.3-70b-instruct), prompt version, criterion scores, and execution timestamp, is immutably logged for complete scholarly reproducibility and transparency.

4. Purposes and Legal Grounds of Processing

Data is processed under the legal grounds of contract performance, statutory obligations, and legitimate scholarly publishing interest (open access indexing, permanent Crossref/DataCite DOI minting, and long-term digital archiving under CC-BY 4.0).

5. Data Subject Rights & Contact

Authors and users may exercise their rights of access, rectification, erasure, and objection to automated decisions by contacting the Editorial Office: [email protected].