The Dark Side of Generative AI: A Cross‐Domain Systematic Investigation
Open Access
- 1 January 2026
- journal article
- Published by Wiley in Human Behavior and Emerging Technologies
- Vol. 2026 (1)
Abstract
Today, generative AI enables the creation of nearly any type of content within seconds at a user′s request. This request is not a cryptic command entered into a dark terminal but rather a prompt written or spoken in natural language via user‐friendly interfaces. This transformative technology has made AI tools widely accessible to the public. As generative AI continues to attract attention and experience rapid global adoption, its darker aspects have become increasingly evident. Despite its numerous advantages, growing concerns related to risks, threats, and ethical challenges demand urgent attention and mitigation. This systematic literature review investigates the dark side of generative AI by analyzing insights from 53 peer‐reviewed articles across 50 scholarly sources, published between 2023 and 2025. Sourced exclusively from the Scopus and Web of Science databases, the reviewed literature spans a variety of domains, including education, healthcare, law, business, and engineering, to explore how ethical issues and technological risks manifest in domain‐specific contexts. The review addresses three core questions: (1) What are the key insights presented in the selected studies? (2) What risks and threats are associated with generative AI, and how do they affect different sectors? (3) What research agenda should guide future studies aimed at addressing these risks and challenges? The findings reveal 21 distinct categories of generative AI′s dark side, including misinformation, bias, privacy violations, academic dishonesty, and job displacement. Through thematic analysis and structured presentation, this review maps the current landscape, identifies research gaps, and outlines future directions. Ultimately, it offers valuable guidance for researchers, educators, developers, and policymakers seeking to mitigate harm while preserving the benefits of innovation.Keywords
This publication has 68 references indexed in Scilit:
- Generative AI for synthetic data across multiple medical modalities: A systematic review of recent developments and challengesComputers in Biology and Medicine, 2025
- Impromptu: a framework for model-driven prompt engineeringSoftware and Systems Modeling, 2025
- ChatGPT for natural sciences course design: Insights from a strengths, weaknesses, opportunities, and threats analysisNatural Sciences Education, 2024
- Moderating Synthetic Content: the Challenge of Generative AIPhilosophy & Technology, 2024
- Addressing the risks of generative AI for the judiciary: The accountability framework(s) under the EU AI ActComputer Law & Security Review, 2024
- Being Human in the Age of Generative AI: Young People's Ethical Concerns about Writing and Living with MachinesReading Research Quarterly, 2024
- Generative AI in higher education and beyondBusiness Horizons, 2024
- Complications Following Body Contouring: Performance Validation of Bard, a Novel AI Large Language Model, in Triaging and Managing Postoperative Patient ConcernsAesthetic Plastic Surgery, 2024
- Chatbots and mental health: Insights into the safety of generative AIJournal of Consumer Psychology, 2023
- ChatGPT in Head and Neck Oncology-Opportunities and ChallengesIndian Journal of Otolaryngology and Head & Neck Surgery, 2023