Algorithmic bias and discriminatory outcomes from machine

The gap

Evidence profile

Sourced from the limitations and open questions and future work and inline gaps of the source papers, classified as methodology gap, spanning 4 journals. Those papers have been cited 4 times in total.

Supporting evidence — 5 representative gaps

• Consensus and legitimation in global AI regulations: a sociosemiotic perspective (2026) · International Journal of Law in Context · cited 4× · doi

  The paper identifies that bias discourse in AI regulatory documents concentrates on bias avoidance rather than articulating concrete operational criteria for unbiased AI systems. Developing specific, measurable definitions of algorithmic bias that account for the spatiotemporal dependency of bias across different legal cultures and regions is needed to move beyond vague modifiers like 'harmful' or 'unjust' toward enforceable standards in global AI regulations.

  methodology gaplimitationsevidence 5/5
  Keywords: algorithmic bias operational criteria global AI regulations semantic uncertainty spatiotemporal dependency
• AI Governance and Ethical Accountability in Indian Corporates: A New Dimension of Corporate Governance (2026) · Iconic Research and Engineering Journals · doi

  The paper identifies that India currently follows a policy-driven approach to AI governance without binding legal rules, but does not specify which existing Indian corporate governance statutes (Companies Act 2013, Information Technology Act 2000) require amendment or what specific regulatory mechanisms should replace voluntary compliance frameworks to enforce AI accountability in corporate decision-making.

  methodology gapopen questionsevidence 5/5
  Keywords: AI governance India regulatory framework Companies Act Information Technology Act corporate accountability enforcement mechanisms
• AI Governance and Ethical Accountability in Indian Corporates: A New Dimension of Corporate Governance (2026) · Iconic Research and Engineering Journals · doi

  The proposed Responsible AI Governance Model emphasizes human oversight, transparency, and ethical accountability in AI-driven corporate decisions, but lacks specification of how board-level committees should operationalize risk-based AI categorization (adapted from EU AI Act) within Indian corporate structures, including audit protocols for algorithmic bias detection in hiring, financial analysis, and customer insights applications.

  methodology gapfuture workevidence 5/5
  Keywords: AI governance model board-level oversight risk-based categorization algorithmic bias audit protocols corporate decision-making
• Desafios regulatórios e éticos relativos ao uso da inteligência artificial na prevenção e repressão à lavagem de dinheiro (2026) · Revista do Tribunal Regional Federal da 3ª Região · doi

  While audit mechanisms for AI algorithms are recommended as essential to regulatory compliance, the paper does not specify what audit standards, metrics, explainability requirements (LIME, SHAP values, attention mechanisms), or validation datasets should be used to audit AML/CFT AI models for bias detection and mitigation before deployment.

  methodology gapinline gapsevidence 5/5
  Keywords: algorithm auditing explainability AML/CFT bias detection model validation machine learning
• Strategic value driven by artificial intelligence in global businesses: a bibliometric and qualitative analysis of the most influential literature (2026) · Frontiers in Artificial Intelligence · doi

  While algorithmic bias and discriminatory outcomes from machine learning models trained on incomplete or historically biased datasets are documented, there is no systematic empirical framework for auditing and measuring bias severity across different industry sectors and organizational contexts in AI-supported decision systems.

  methodology gapopen questionsevidence 5/5
  Keywords: algorithmic bias machine learning models discriminatory outcomes audit mechanisms bias measurement AI governance

Questions about this gap