From Data Protection to Algorithmic Accountability: Rethinking Legal Remedies for AI-Assisted Decisions in Ghana
Public bodies and private institutions in Ghana increasingly rely on artificial intelligence to allocate credit, screen applicants, select taxpayers for audit and determine eligibility for services, yet the statutes that govern these decisions were enacted for an earlier technological order. This article contends that the resulting deficit is chiefly remedial. The 1992 Constitution guarantees privacy, equality, access to information and administrative justice, and the Data Protection Act, 2012 (Act 843) contains an early African provision on decisions based solely on automated processing; but these instruments do not reliably secure for an affected person the conditions on which an effective remedy depends. Those conditions are notice that an automated system contributed to the decision, reasons sufficient to permit a challenge, a record from which the decision can be reconstructed, review by a person empowered to alter the outcome, and redress proportionate to the harm. The article develops these conditions as the Remedial Chain Test and applies it to the law in force at 29 October 2026. It distinguishes binding law from the proposals in the Data Protection Bill, 2025, the Cybersecurity (Amendment) Bill, 2025 and the Emerging Technologies Bill, and it identifies the evidence that regulators, courts and researchers would require to establish whether the framework operates in practice.
Authors
- Godwin Gaduga
Institutions
- Accra Institute of Technology (GH)
- BlueCrest University College (GH)
Publication Details
- Journal
- International Journal of innovative inventions in Social Science and Humanities
- Published
- 2026-11-20
- DOI
- https://doi.org/10.58806/ijiissh.2026.v3i11n06
- Primary Topic
- Digitalization, Law, and Regulation
- Type
- article
- Field-Weighted Citation Impact
- 0.00