Nigeria And UK Build Public-Sector Capacity For AI Standards

Nigeria and the United Kingdom have completed a four-day executive programme intended to help public officials apply international standards to artificial-intelligence management and risk.

The programme was delivered by the British Standards Institution and Nigeria's National Centre for Artificial Intelligence and Robotics under the UK's Digital Access Programme. BusinessDay reported that participants came from the federal communications ministry, NITDA, the Standards Organisation of Nigeria, the Nigeria Data Protection Commission and NCAIR.

The curriculum covered ISO/IEC 42001, the international standard for an AI management system, and ISO/IEC 23894, guidance on AI risk management. These frameworks can help an institution define responsibility, assess risk, maintain documentation and review systems across their lifecycle. They are not, by themselves, laws or proof that an AI product is safe.

That distinction is important. The event strengthened capacity; it did not enact a new Nigerian regulation or certify all participating agencies. Its value will depend on what officials change in procurement, project approval and oversight after returning to their institutions.

Nigeria's public sector is already encountering AI through document tools, biometric systems, service chatbots, risk analysis and security proposals. Agencies need a common method for asking basic questions: What is the intended decision? Which data is used? Who is accountable for errors? How can an affected person seek review? What evidence is required before a system is expanded?

ISO-based processes can provide structure, particularly when ministries buy technology from vendors with stronger technical knowledge than the contracting authority. A tender can require model documentation, evaluation data, incident reporting, audit access and exit arrangements. Without these terms, a public body may discover too late that it cannot inspect a system or move its data.

NCAIR national director Olubunmi Ajala called for a permanent cross-government community of practice on AI governance. That would be a practical next step. Officials dealing with standards, data protection, digital policy and sector regulation often examine different parts of the same deployment. A shared forum can identify conflicts early and prevent agencies from reinventing controls.

The group should publish usable material rather than keeping the training inside a small cohort. A model risk register, procurement clauses, impact-assessment template and incident protocol would help ministries apply the principles consistently. Public guidance would also allow companies and civil society to understand the standard expected of government systems.

International standards still need Nigerian interpretation. Local languages, uneven record quality, connectivity constraints and informal service channels affect performance. A technically accurate system may exclude citizens without a digital identity or stable device. Risk assessment must therefore cover access and administrative justice, not only cybersecurity and model metrics.

BSI has recently expanded its AI testing and assurance services, including work related to management systems and product risk. Nigeria should preserve institutional independence when drawing on that expertise. Standards bodies can advise and assess, while Nigerian regulators and elected authorities must decide what public accountability requires.

The programme is a useful investment in state capability. The measure of success will not be the number of officials trained, but whether public AI purchases become easier to inspect, challenge and improve. Nigeria now needs to turn a four-day course into routine governance that follows every system from proposal to retirement.

Sources: BusinessDay, NCAIR policy and ethics programme and BSI on AI management and risk assurance.