Eswatini Aice Africa AI Safety Centre Established by ESWASA

• The Eswatini Standards Authority (ESWASA) and AICE Africa have launched an AI Safety and Standards Centre of Excellence in Mbabane.
• The centre aims to build Africa's capacity for AI system assessment, governance, and certification, reducing reliance on external institutions.
• It will develop technical skills, support AI adoption, and establish a research lab to evaluate AI against local and international standards.
• The initiative plans to utilize ISO/IEC 42001 for AI management systems and ISO/IEC 23894 for AI risk management.
• Future plans include professional training, industry services, regional cooperation, and exploring certification services.

True AI sovereignty necessitates the development of internal capabilities to evaluate, govern, and enhance the systems upon which societies rely.

The Eswatini Standards Authority (ESWASA) and AICE Africa have jointly established an Artificial Intelligence (AI) Safety and Standards Centre of Excellence, a significant initiative launched in Mbabane on October 27. This new Eswatini Aice Africa AI Safety Centre is designed to bolster Africa's capabilities in the assessment, governance, and certification of AI systems across the continent. Its primary objective is to cultivate technical expertise, assist both private enterprises and public institutions in their adoption of AI technologies, and develop a dedicated research and testing laboratory.

This proposed laboratory will be crucial for evaluating AI systems, ensuring they align with local conditions, including African languages, datasets, infrastructure, and social contexts, alongside recognized international standards. The establishment of this centre comes at a critical juncture, as governments and businesses increasingly integrate AI into their operations, leading to heightened concerns regarding data protection, system reliability, accountability, and the inherent risks associated with automated decision-making processes. The ESWASA AICE Africa partnership aims to address these challenges head-on by fostering robust AI governance standards.

Under the terms of this collaborative effort, ESWASA will provide essential institutional leadership, leverage its extensive standards expertise, and coordinate national efforts. Concurrently, AICE Africa will contribute its specialized knowledge in AI governance, oversee technical implementation, drive program development, and facilitate crucial skills transfer. Ncamiso Mhlanga, Executive Director of ESWASA, emphasized the necessity for standards bodies to effectively evaluate AI as it becomes integral to business and public services, noting that the organization is undergoing an internal transformation to build the necessary human capacity.

Initially, the Eswatini Aice Africa AI Safety Centre will concentrate on enhancing ESWASA's internal operational efficiency before expanding its services to include professional training, industry support, and broader regional cooperation. A key technological component of this initiative is EswasaOne, an AI-powered platform currently under development. This platform is envisioned to integrate standards information, streamline certification and inspection workflows, and manage institutional processes, featuring a conversational interface and automated workflows to improve accessibility to standards data and optimize internal operations.

A central goal of the Eswatini Aice Africa AI Safety Centre is to reduce Africa's dependence on external entities for the evaluation of AI technologies deployed within the continent. John Kamara, CEO of AICE Africa, articulated that true AI sovereignty necessitates the development of internal capabilities to evaluate, govern, and enhance the systems upon which societies rely. The centre plans to draw extensively from international benchmarks, specifically ISO/IEC 42001, which outlines requirements for AI management systems, and ISO/IEC 23894, offering comprehensive guidance on AI risk management Eswatini. This commitment to international standards is pivotal for establishing credible Eswatini AI governance standards.

Furthermore, the initiative includes plans for developing specialized practitioner training programs tailored for standards officers, auditors, regulators, and technical personnel. Universities and other training institutions are expected to play a supportive role in both research and professional development. While the centre intends to explore the provision of certification services, the realization of an African AI certification centre will hinge on the development of validated assessment methodologies, the availability of qualified personnel, and appropriate accreditation arrangements. To further regional cooperation, the partners are planning to host an AI Safety and Standards Summit in Mbabane in July 2027.

Lawyers and compliance officers should monitor the development of AI safety and standards in Eswatini, particularly as the new centre aims to establish certification services and training based on international standards like ISO/IEC 42001. This initiative signals future regulatory and compliance requirements for AI systems operating within or impacting Eswatini, necessitating proactive assessment of AI deployments for adherence to emerging governance frameworks.

Source: Original reporting via Capital FM