This article was automatically translated from HIBAPRESS, the Arabic version:
Hibapress-rabat-mef AI, often described as fourth industrial revolution, profoundly transforms all sectors, and public service is no exception to this dynamic. Well beyond the automation of repetitive tasks, AI offers unprecedented capacities in terms of predictive analysis, identification of anomalies, detection of fraud and process optimization. In the field of public internal audit, these advances mark the emergence of the increased audit, where human intelligence collaborates closely with powerful algorithms to increase the efficiency and precision of missions. This rise in power of algorithms is revolutionizing the practice of audit. Audit algorithms, capable of processing massive volumes of real -time data, redefine the traditional role of the listener. The latter could gradually become the machine listener, responsible for verifying the relevance of models, data integrity and the absence of biases in algorithmic decisions.
Thus, the integration of AI into the public internal audit is not limited to a simple technological evolution, but constitutes a fundamental mutation. It requires institutions to adopt a proactive approach to support this transformation, forming auditors to new tools, by adapting regulatory frameworks, and by anticipating the ethical and technical issues linked to this man-machine cohabitation. This transition, if it is well controlled, opens the way to more transparent, resilient and innovative public governance. However, this dynamic is accompanied by fundamental questions, which call for rigorous and enlightened governance, in order to ensure that AI becomes a lever for transparency and equity, rather than a factor of amplification of inequalities or of confidentiality. Its integration into the public internal audit, although offering promising perspectives, is not free from obstacles. Among the major challenges that are loomed in particular: • Algorithmic biases and equity: AI algorithms, although powerful, are often influenced by the biases present in the data which feeds them and the arbitrations they integrate. In the public context, these biases can impact the relevance of decisions and alter confidence in institutions. • Data processing: use of AI requires access to large amounts of data, often sensitive. This raises questions about the confidentiality and security of the information processed, especially in sensitive fields and also the challenges of development of physical infrastructure (servers, networks, security, energy, etc.). • Skills deficit: artificial intelligence requires specific technical skills, such as mastery of data analysis tools, programming languages, and an in -depth understanding of algorithmic models. These skills are still lacking in many audit bodies and it is imperative that listeners in the future are trained in the use of AI analysis tools, in order to act as agile and innovative actors. • unsuitable legal and normative frameworks: current legislation and regulation do not always cover the specificities of AI, leaving risk areas as to its use and legal implications.
