ai-research-survey

scan.json (24554B)

---

 {
   "paper": {
     "title": "Systematic Literature Review on Generative AI: Ethical Challenges and Opportunities",
     "authors": ["Feliks Prasepta Sejahtera Surbakti"],
     "year": 2025,
     "venue": "International Journal of Advanced Computer Science and Applications (IJACSA)",
     "doi": "10.14569/ijacsa.2025.0160530"
   },
   "scan_version": 3,
   "active_modules": ["survey_methodology"],
   "methodology_tags": ["meta-analysis"],
   "key_findings": "This SLR of 112 peer-reviewed articles identifies five core ethical challenges in generative AI: algorithmic bias, misinformation/deepfakes, data privacy violations, intellectual property issues, and accountability/explainability. The review finds that GAI ethics research is heavily fragmented across disciplines with inconsistent terminology and limited empirical validation. The paper also identifies opportunities for ethical innovation through responsible design, inclusive governance, and interdisciplinary collaboration, and notes a dramatic surge in publications in 2024 (58 articles).",
   "checklist": {
     "artifacts": {
       "code_released": {
         "applies": true,
         "answer": false,
         "justification": "No analysis code or scripts are released. A survey can release its analysis code (e.g., coding scripts, bibliometric tools), but none is provided."
       },
       "data_released": {
         "applies": true,
         "answer": false,
         "justification": "The full list of 112 reviewed articles is not provided as a downloadable dataset. Table I shows only 10 of 112 entries. No supplementary materials or data files are referenced."
       },
       "environment_specified": {
         "applies": true,
         "answer": false,
         "justification": "No computational environment is specified. If any software tools were used for the thematic analysis or bibliometric analysis, they are not mentioned."
       },
       "reproduction_instructions": {
         "applies": true,
         "answer": false,
         "justification": "While the search strategy and PRISMA flow are described, there are no step-by-step reproduction instructions. The search query is provided but the full list of included papers is not, making independent reproduction impossible."
       }
     },
     "statistical_methodology": {
       "confidence_intervals_or_error_bars": {
         "applies": false,
         "answer": false,
         "justification": "This is a thematic literature review with no statistical analysis. No quantitative experiments are performed that would require confidence intervals."
       },
       "significance_tests": {
         "applies": false,
         "answer": false,
         "justification": "The paper performs thematic synthesis, not statistical analysis. No comparative statistical claims are made."
       },
       "effect_sizes_reported": {
         "applies": false,
         "answer": false,
         "justification": "No experimental comparisons are made. The paper reports article counts by category but these are descriptive, not effect sizes."
       },
       "sample_size_justified": {
         "applies": false,
         "answer": false,
         "justification": "As a thematic literature review, sample size justification in the experimental sense is not applicable. The corpus size of 112 is a result of the search/filter process, not a design choice."
       },
       "variance_reported": {
         "applies": false,
         "answer": false,
         "justification": "No experimental runs or repeated measurements are conducted. This is a qualitative thematic synthesis."
       }
     },
     "evaluation_design": {
       "baselines_included": {
         "applies": true,
         "answer": false,
         "justification": "Section II.D discusses related systematic and scoping reviews (Hagendorff 2024, Tabassum et al. 2025, Al-kfairy et al. 2024, Zhang & Boulos 2023) but does not systematically compare its findings, coverage, or methodology against them. They are positioned as related work, not as baselines for comparison."
       },
       "baselines_contemporary": {
         "applies": true,
         "answer": false,
         "justification": "No formal baseline comparison is conducted, so contemporaneity of baselines cannot be assessed."
       },
       "ablation_study": {
         "applies": false,
         "answer": false,
         "justification": "This is a literature review with no system components to ablate."
       },
       "multiple_metrics": {
         "applies": false,
         "answer": false,
         "justification": "No experimental evaluation metrics are used. The paper performs qualitative thematic analysis."
       },
       "human_evaluation": {
         "applies": false,
         "answer": false,
         "justification": "No system outputs to evaluate. This is a literature review."
       },
       "held_out_test_set": {
         "applies": false,
         "answer": false,
         "justification": "No experimental test sets are used. This is a qualitative literature review."
       },
       "per_category_breakdown": {
         "applies": true,
         "answer": true,
         "justification": "Findings are broken down by ethical challenge category (bias, misinformation, privacy, IP, accountability), by domain (healthcare, education, media, law), by publication year (Fig. 1), by subject area (Fig. 2), by journal (Fig. 3), and by article type (Fig. 4)."
       },
       "failure_cases_discussed": {
         "applies": true,
         "answer": true,
         "justification": "The paper identifies gaps and weaknesses in the reviewed literature, including fragmented approaches, lack of empirical grounding, and disproportionate focus on risks over solutions (Section V.C). This functions as discussion of 'failure cases' in the surveyed field."
       },
       "negative_results_reported": {
         "applies": true,
         "answer": true,
         "justification": "The paper reports several negative findings: the research landscape is fragmented, most studies lack empirical validation, current frameworks are inconsistent, and the literature is heavily skewed toward risk identification rather than solutions (Sections IV.D, V.C)."
       }
     },
     "claims_and_evidence": {
       "abstract_claims_supported": {
         "applies": true,
         "answer": true,
         "justification": "The abstract's main claims — five primary ethical challenges, fragmented landscape, limited empirical validation, and emerging opportunities — are all supported by the thematic analysis presented in Sections IV and V."
       },
       "causal_claims_justified": {
         "applies": false,
         "answer": false,
         "justification": "The paper makes no causal claims. It synthesizes and categorizes existing literature thematically without claiming causal relationships between variables."
       },
       "generalization_bounded": {
         "applies": true,
         "answer": false,
         "justification": "The title claims a comprehensive 'Systematic Literature Review on Generative AI: Ethical Challenges and Opportunities' but the search was limited to a single database (ScienceDirect). The paper acknowledges ScienceDirect as a limitation in one sentence ('While recognizing that ethical discussions on GAI may also exist in other databases') but still makes broad claims about the state of the field without adequately bounding them to this single-source scope."
       },
       "alternative_explanations_discussed": {
         "applies": true,
         "answer": false,
         "justification": "The paper does not consider alternative interpretations of its thematic findings. For example, the observed fragmentation could reflect genuine disciplinary diversity rather than a problem, but this is not explored."
       },
       "proxy_outcome_distinction": {
         "applies": true,
         "answer": true,
         "justification": "The paper's claims match the granularity of its analysis. It claims to identify themes in the literature and presents those themes. No proxy gap exists between what was measured (thematic categories in 112 papers) and what is claimed."
       }
     },
     "setup_transparency": {
       "model_versions_specified": {
         "applies": false,
         "answer": false,
         "justification": "No AI models are used in the study methodology. This is a manual literature review."
       },
       "prompts_provided": {
         "applies": false,
         "answer": false,
         "justification": "No prompting is used. This is a manual literature review."
       },
       "hyperparameters_reported": {
         "applies": false,
         "answer": false,
         "justification": "No computational hyperparameters are relevant to this manual literature review."
       },
       "scaffolding_described": {
         "applies": false,
         "answer": false,
         "justification": "No agentic scaffolding is used. This is a manual literature review."
       },
       "data_preprocessing_documented": {
         "applies": true,
         "answer": true,
         "justification": "The paper documents the PRISMA-based selection pipeline with counts and criteria: 145 initial articles → 12 non-peer-reviewed excluded → 133 screened → 9 removed (not meeting inclusion criteria) → 124 full-text assessed → 6 excluded (insufficient ethical focus) → 112 final. Inclusion/exclusion criteria are stated in Section III.E."
       }
     },
     "limitations_and_scope": {
       "limitations_section_present": {
         "applies": true,
         "answer": false,
         "justification": "There is no dedicated limitations section discussing the study's own methodological limitations. The paper discusses gaps and limitations in the reviewed literature (Sections II.E, V.C) but never acknowledges its own limitations such as the single-database search strategy."
       },
       "threats_to_validity_specific": {
         "applies": true,
         "answer": false,
         "justification": "No threats to the validity of this study's own methodology are discussed. The paper does not address risks such as selection bias from using only ScienceDirect, potential subjectivity in thematic coding, or the impact of the narrow search terms."
       },
       "scope_boundaries_stated": {
         "applies": true,
         "answer": false,
         "justification": "While the search scope is defined (ScienceDirect, 2021-2025, English, peer-reviewed), the paper does not explicitly state what its findings do NOT show. The conclusions make broad claims about the GAI ethics landscape without bounding them to the single-database, single-search-string scope."
       }
     },
     "data_integrity": {
       "raw_data_available": {
         "applies": true,
         "answer": false,
         "justification": "Only 10 of 112 reviewed articles are shown in Table I. The full dataset of extracted data (bibliographic details, ethical themes, quality scores) is not available for independent verification."
       },
       "data_collection_described": {
         "applies": true,
         "answer": true,
         "justification": "The data collection procedure is described in Section III.D: ScienceDirect search using Boolean operators combining 'generative AI', 'ethics', 'challenges', and 'opportunities', limited to peer-reviewed journals in English from 2021-2025."
       },
       "recruitment_methods_described": {
         "applies": true,
         "answer": true,
         "justification": "The paper selection process is described: structured search on ScienceDirect with specific Boolean queries, followed by PRISMA-guided screening at three stages (identification, screening, eligibility). The search strategy and selection criteria are documented."
       },
       "data_pipeline_documented": {
         "applies": true,
         "answer": true,
         "justification": "The PRISMA flow is documented with counts at each stage: 145 → 133 (12 non-peer-reviewed removed) → 124 (9 not meeting criteria removed) → 112 (6 insufficient ethical focus removed). Data extraction form contents and dual-reviewer process are described in Section III.G."
       }
     },
     "conflicts_of_interest": {
       "funding_disclosed": {
         "applies": true,
         "answer": false,
         "justification": "No funding information is provided anywhere in the paper. There is no acknowledgments section mentioning grants or sponsors."
       },
       "affiliations_disclosed": {
         "applies": true,
         "answer": true,
         "justification": "The author's affiliation is clearly listed: Industrial Engineering Department, Atma Jaya Catholic University of Indonesia, Jakarta, Indonesia."
       },
       "funder_independent_of_outcome": {
         "applies": true,
         "answer": false,
         "justification": "No funding is disclosed, so independence of funding cannot be verified."
       },
       "financial_interests_declared": {
         "applies": true,
         "answer": false,
         "justification": "No competing interests or financial interests statement is present in the paper."
       }
     },
     "contamination": {
       "training_cutoff_stated": {
         "applies": false,
         "answer": false,
         "justification": "This is a literature review that does not evaluate any pre-trained model's capability on a benchmark."
       },
       "train_test_overlap_discussed": {
         "applies": false,
         "answer": false,
         "justification": "This is a literature review that does not evaluate any pre-trained model on a benchmark."
       },
       "benchmark_contamination_addressed": {
         "applies": false,
         "answer": false,
         "justification": "This is a literature review that does not evaluate any pre-trained model on a benchmark."
       }
     },
     "human_studies": {
       "pre_registered": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study. This is a systematic literature review."
       },
       "irb_or_ethics_approval": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       },
       "demographics_reported": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       },
       "inclusion_exclusion_criteria": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       },
       "randomization_described": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       },
       "blinding_described": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       },
       "attrition_reported": {
         "applies": false,
         "answer": false,
         "justification": "No human participants in this study."
       }
     },
     "cost_and_practicality": {
       "inference_cost_reported": {
         "applies": false,
         "answer": false,
         "justification": "This is a survey paper with no computational method to cost."
       },
       "compute_budget_stated": {
         "applies": false,
         "answer": false,
         "justification": "This is a survey paper with no computational budget."
       }
     },
     "survey_methodology": {
       "prisma_or_structured_protocol": {
         "applies": true,
         "answer": true,
         "justification": "The paper explicitly follows PRISMA 2020 guidelines (citing Page et al. 2021) and uses a structured SLR protocol based on Kitchenham et al. (2007) and Okoli (2015). The PRISMA flow with stage counts is described in Section III.F."
       },
       "quality_assessment_of_sources": {
         "applies": true,
         "answer": true,
         "justification": "A CASP-based quality assessment checklist is used with five criteria (clarity of objective, relevance to GAI ethics, methodological soundness, contribution, transparency of limitations), each scored 0-1, with a minimum threshold of 3/5 for inclusion. Two reviewers independently assessed all articles (Section III.H)."
       },
       "publication_bias_discussed": {
         "applies": true,
         "answer": false,
         "justification": "The paper does not discuss publication bias. There are no funnel plots, no discussion of negative-result underrepresentation, and no acknowledgment that published GAI ethics papers may skew toward identifying risks over reporting neutral or positive findings."
       }
     }
   },
   "claims": [
     {
       "claim": "Five primary ethical challenges dominate GAI discourse: algorithmic bias, misinformation/deepfakes, data privacy violations, intellectual property issues, and accountability/explainability.",
       "evidence": "Identified through thematic analysis of 112 peer-reviewed articles using open and axial coding (Section IV.B). Each challenge is discussed with citations to supporting studies.",
       "supported": "moderate"
     },
     {
       "claim": "The GAI ethics research landscape is fragmented across disciplines with limited empirical validation and inconsistent ethical frameworks.",
       "evidence": "Section V.C discusses cross-disciplinary fragmentation, citing Hagendorff (2024) who 'concluded that terminological and methodological fragmentation hinders ethical progress.' The paper notes 'most articles focus disproportionately on ethical risks' and there is 'a notable scarcity of empirical research.'",
       "supported": "moderate"
     },
     {
       "claim": "Publication interest in GAI ethics surged dramatically in 2024, with 58 articles published.",
       "evidence": "Bibliometric analysis presented in Fig. 1 shows: 2022 (n=1), 2023 (n=5), 2024 (n=58), 2025 (n=48). However, these counts are from a single database (ScienceDirect) with specific search terms.",
       "supported": "moderate"
     },
     {
       "claim": "GAI can be ethically harnessed to promote inclusive innovation, assistive technologies, and human-AI creative collaboration.",
       "evidence": "Section IV.C and V.B cite studies on democratizing creativity, assistive technologies for people with disabilities (Elyoseph et al. 2024), and human-centered design principles. However, these are drawn from the reviewed literature, not original empirical findings.",
       "supported": "weak"
     }
   ],
   "red_flags": [
     {
       "flag": "Single database search",
       "detail": "The review searched only ScienceDirect, excluding major sources like Scopus, Web of Science, IEEE Xplore, ACM Digital Library, PubMed, and arXiv. For a 'systematic' review claiming to map the ethical landscape of GAI, this severely limits coverage and introduces selection bias. The paper acknowledges this briefly but does not treat it as a significant limitation."
     },
     {
       "flag": "No self-limitations section",
       "detail": "The paper discusses gaps in the reviewed literature extensively but never acknowledges the methodological limitations of its own study. A systematic review that doesn't discuss its own threats to validity (single database, narrow search terms, potential coding subjectivity) is a significant omission."
     },
     {
       "flag": "Inconsistent date range",
       "detail": "The abstract and methodology state the review period is '2021 and 2025,' but the conclusion states 'between 2018 and early 2025.' This inconsistency undermines confidence in the review's rigor."
     },
     {
       "flag": "Incomplete data presentation",
       "detail": "Table I shows only 10 of 112 reviewed articles. Without the full dataset, the thematic findings cannot be independently verified. The quality assessment scores are never reported, making it impossible to assess the actual quality distribution of included studies."
     },
     {
       "flag": "Dual-reviewer claim without evidence",
       "detail": "The paper claims 'two reviewers independently extracted data' and independently assessed quality (Section III.G-H), but this is a single-author paper. The identity of the second and third reviewers is never disclosed, and no inter-rater reliability statistics are reported."
     },
     {
       "flag": "Shallow thematic synthesis",
       "detail": "The five identified ethical themes (bias, misinformation, privacy, IP, accountability) are well-established in the literature and could be identified without a systematic review. The paper adds little novel synthesis beyond cataloging known concerns, which is the 'laundering signal-to-noise ratio' problem for surveys."
     }
   ],
   "cited_papers": [
     {
       "title": "Language models are few-shot learners",
       "authors": ["T. Brown", "B. Mann", "N. Ryder"],
       "year": 2020,
       "relevance": "Foundational GPT-3 paper relevant to understanding large language model capabilities that underpin generative AI."
     },
     {
       "title": "Attention is all you need",
       "authors": ["A. Vaswani", "N. Shazeer", "N. Parmar"],
       "year": 2017,
       "relevance": "Foundational Transformer architecture paper underlying all modern generative AI models."
     },
     {
       "title": "On the dangers of stochastic parrots: Can language models be too big?",
       "authors": ["E. M. Bender", "T. Gebru", "A. McMillan-Major", "S. Shmitchell"],
       "relevance": "Seminal paper on ethical risks of large language models including bias, environmental costs, and misleading outputs."
     },
     {
       "title": "A comprehensive empirical study of bias mitigation methods for machine learning classifiers",
       "authors": ["Z. Chen", "J. M. Zhang", "F. Sarro", "M. Harman"],
       "year": 2023,
       "relevance": "Empirical study of bias mitigation methods for ML classifiers, directly relevant to AI fairness evaluation."
     },
     {
       "title": "Towards A Rigorous Science of Interpretable Machine Learning",
       "authors": ["F. Doshi-Velez", "B. Kim"],
       "year": 2017,
       "relevance": "Foundational paper on AI explainability and interpretability, relevant to accountability and transparency in AI systems."
     },
     {
       "title": "Mapping the ethics of generative ai: A comprehensive scoping review",
       "authors": ["T. Hagendorff"],
       "year": 2024,
       "relevance": "Closely related scoping review categorizing GAI ethical concerns into nineteen thematic areas, directly comparable to this survey."
     },
     {
       "title": "ChatGPT: A comprehensive review on background, applications, key challenges, bias, ethics, limitations and future scope",
       "authors": ["P. P. Ray"],
       "year": 2023,
       "relevance": "Comprehensive review of ChatGPT covering capabilities, ethics, and limitations relevant to LLM evaluation and safety."
     },
     {
       "title": "Ethical challenges and solutions of generative AI: An interdisciplinary perspective",
       "authors": ["M. Al-kfairy", "D. Mustafa", "N. Kshetri", "M. Insiew", "O. Alfandi"],
       "year": 2024,
       "relevance": "Cross-sectoral review of GAI ethical challenges across industries, directly relevant to AI governance and responsible development."
     },
     {
       "title": "Generative AI and the Metaverse: A Scoping Review of Ethical and Legal Challenges",
       "authors": ["A. Tabassum", "E. Elmahjub", "A. I. Padela", "A. Zwitter", "J. Qadir"],
       "year": 2025,
       "relevance": "Scoping review of GAI ethical and legal challenges in metaverse applications including bias, disinformation, and privacy."
     }
   ],
   "engagement_factors": {
     "practical_relevance": {
       "score": 0,
       "justification": "Pure literature review with no actionable tool, technique, or framework that practitioners could apply."
     },
     "surprise_contrarian": {
       "score": 0,
       "justification": "Findings confirm well-established ethical concerns about generative AI (bias, privacy, misinformation) without challenging any conventional wisdom."
     },
     "fear_safety": {
       "score": 1,
       "justification": "Discusses AI safety concerns like deepfakes and misinformation but presents no novel threats or demonstrations beyond what is already widely known."
     },
     "drama_conflict": {
       "score": 0,
       "justification": "No controversy, no challenge to specific actors, no dramatic framing."
     },
     "demo_ability": {
       "score": 0,
       "justification": "No code, demo, or interactive component."
     },
     "brand_recognition": {
       "score": 0,
       "justification": "Published in IJACSA by a solo author at Atma Jaya Catholic University of Indonesia; no major lab or product involvement."
     }
   }
 }