Africa seeks control of AI safety as Eswatini launches standards centre

Africa is seeking a greater role in determining how artificial intelligence (AI) technologies are developed, tested and regulated, amid concerns that the continent remains dependent on foreign expertise to assess systems increasingly shaping its economies and public services.

“AI sovereignty requires institutions that can evaluate, govern and improve the systems they depend on,” said John Kamara, Chief Executive Officer of AICE Africa.

The remarks came during the launch of an AI Safety and Standards Centre of Excellence in Eswatini on October 27, an initiative aimed at building African expertise in AI governance, technical assessment and safety standards.

The centre, established through a partnership between AICE Africa and the Eswatini Standards Authority (ESWASA), will train professionals, strengthen institutional capacity and develop a proposed AI research and testing laboratory to evaluate technologies against African realities.

“As AI becomes part of how businesses and public services operate, standards bodies must be able to assess it,” said ESWASA Executive Director Ncamiso Mhlanga.

He said the authority would begin by transforming its internal operations before extending its expertise to businesses and other institutions across the continent.

The initiative seeks to address Africa's limited capacity to independently determine whether AI systems are safe, fair and suitable for local populations, particularly as their use expands in healthcare, education, agriculture and financial services.

The proposed laboratory will develop shared assessment methods, benchmarking resources and an African AI Standards Baseline Report to guide regulators, businesses and standards institutions.

The partners will also draw on international frameworks, including ISO/IEC 42001 on AI management systems and ISO/IEC 23894 on AI risk management, while seeking to incorporate African experiences into global standards development.

The partnership will see ESWASA provide institutional leadership and standards expertise, while AICE Africa contributes technical implementation, AI governance and skills transfer.

A training programme will target standards officers, auditors, regulators and technical personnel, with universities expected to support academic programmes and applied research.

The centre also aims to help African businesses adopt AI responsibly, strengthen consumer protection and reduce dependence on external assessments.

The partners plan to convene an AI Safety and Standards Summit in Mbabane in July 2027 to advance continental cooperation.

Its broader success will depend on whether the initiative translates its ambitions into sustainable African testing capacity and institutions capable of shaping global AI governance.