In 2024, global investment in artificial intelligence exceeded 150 billion dollars. More than two thirds of that amount was concentrated in three regions: the United States, China and the European Union. The capacity to train advanced models requires very high-performance infrastructure, high energy consumption and specialised scientific teams. A single large-scale model may require thousands of processors running for weeks. These figures help us understand an essential point: algorithmic governance begins with infrastructure.
When we talk about rules for artificial intelligence, we often think of laws and regulations. However, before the law comes technical design. Every system embodies choices: which data to use, which objectives to optimise, which criteria to treat as priorities. These decisions are made by engineering teams, international consortia, standard-setting bodies and, increasingly, multilateral forums.
Legitimacy, in this context, is built differently from traditional politics. It results from technical competence, transparency of processes and adherence to internationally recognised standards. Organisations such as the ISO, the OECD and the United Nations have been establishing guiding principles for the responsible development of automated systems. At the same time, regional blocs are adopting their own regulatory frameworks, seeking to balance innovation with the protection of rights.
The geography of technological capacity
The available data show that more than 70% of the world's cloud computing capacity is concentrated in a small number of global companies. Most high-performance data centres are located in North America, Europe and East Asia. This concentration does not imply exclusion, but reflects decades of continuous investment in science, energy and connectivity.
For African countries, including Mozambique, the challenge lies in integrating into this ecosystem strategically. The continent accounts for only about 3% of the world's data centres, but has seen significant growth over the past five years. New infrastructure in South Africa, Kenya, Nigeria and Morocco is expanding regional capacity and reducing external dependence.
The asymmetry in access to advanced computing is real, but it is evolving. International cooperation programmes allow African universities to take part in global research networks. Open-source platforms are democratising tools that were once restricted to large laboratories. Knowledge circulates more freely than at any other moment in history.
Taking part in writing the rules
Algorithmic governance is not a closed space. It is a process under construction. In 2023, more than 50 countries had already published national strategies for artificial intelligence. Many include goals for developing local capacity, incentives for innovation and ethical principles adapted to their own reality.
Legitimacy is born when different regions take an active part in defining standards. International conferences, technical groups and regional forums become spaces where diverse voices help to shape global guidelines. This is not about contesting hegemonies, but about ensuring plurality.
Mozambique, part of a dynamic regional community, can benefit from joint initiatives within SADC and the African Union. Investing in scientific training, encouraging university partnerships and strengthening regulatory capacity are steps that increase the country's presence in this debate.
Between interdependence and opportunity
The digital economy is, by nature, interconnected. Data crosses borders in milliseconds. Services are delivered from different continents. This interdependence does not eliminate sovereignty, but redefines it. Autonomy in the twenty-first century is also measured by the capacity to understand, negotiate and adapt global technologies to national priorities. The risk of asymmetry exists when participation is limited. The opportunity arises when there is strategy. Countries that invest in digital literacy, applied research and adequate infrastructure position themselves as active players, not merely users.
The writing of algorithmic rules is a collective process and one in permanent evolution. It includes engineers, legislators, academics and international organisations. Legitimacy is built on cooperation, transparency and the inclusion of different perspectives.
Technological transformation is not a zero-sum game. It is a phase of global reorganisation, in which knowledge, data and computing capacity become pillars of development. The central question is not who controls the world's brain, but how to ensure that this brain is plural, regulated and oriented towards the common good.
Opening this debate with calm and strategic vision is essential. Because, in the digital age, taking part in defining the rules is also taking part in defining the future.
