Today, personal data and cutting-edge AI-based scoring models play an ever-increasing role in determining who is accepted or rejected for consumer credit. This process is becoming more automated than ever before. Determining whether consumers are creditworthy or not is often obscured within a maze of tenuous transparency. Black Box models pose a serious threat, particularly when a credit applicant is categorized as ‘bad’ payer. This study employs the Diverse Counterfactual Explanations (DiCE) technique in three Use Cases through a Multi-Layer Perceptron (MLP) embedded with the separated combination of J48 Algorithm and Random Forest (RF) for feature selection. The classifier was deliberately pre-processed on the UCI German Credit Dataset (GCD) to clarify the negative impact software engineers have on both scoring and post-hoc explanations. Our findings indicate that while Counterfactuals (CFs) can, in some circumstances, help consumers whose credit was denied, they are often incomplete for providing legally and reasonably grounded explanations of why and how. So, besides imposing a ‘de-individualized’ identity on consumers, CFs do not fulfill all criteria to achieve a grounded right to ex-post justification. We thus inferred that the field of eXplainable AI (XAI) has yet a long path to deliver reasonable solutions to address the challenges posed by EU Law. To date, justifications should be provided through Machine-Scoring-to-Human-Justifications (MS2HJ) Interaction. However, as XAI becomes more and more trustworthy, we do not deny fully automated justifications, embedding legal values and norms (i.e., Humanichal). Until now, DiCE shortcomings suggest that the Portuguese legislator, when transposing Art. 18 of the Directive (EU) 2023,2225, of October 18, should enshrine the right to Human-Centric justifiability based on proportional and legally grounded or actionable alternatives.