Mozambique’s New AI Commission Faces Its First Big Test: Turning Strategy into Action

Mozambique is moving to establish the institutional foundations for artificial intelligence governance. On 17 April 2026, the Council of Ministers approved Decree No. 14/2026, creating the National Commission for Artificial Intelligence (CNIA), a consultative and technical advisory body tasked with helping the government shape the country’s AI policies and regulatory direction.

The commission’s mandate extends beyond ethical guidance. It includes reviewing AI-related policies and legislative proposals, assessing legal and security risks, advising on strategic financing, promoting research and innovation, and examining AI’s potential economic and social impacts. Its membership draws on government, regulatory authorities, academia, civil society and the private sector, with the National Institute of Information and Communication Technologies (INTIC) providing its executive secretariat.

The institutional design reflects a recognition that AI cannot be governed by technology specialists alone. Decisions about automated public services, financial systems, education and critical infrastructure affect citizens’ rights, public spending, employment and national security.

Yet the CNIA’s creation should not be mistaken for the arrival of a comprehensive AI regulatory regime. Its role is primarily advisory: it can help shape government decisions, but its effectiveness will depend on whether its recommendations lead to adopted policies, clear legal responsibilities and adequately resourced implementation.

That distinction is central to understanding Mozambique’s next phase of AI development.

Mozambique presented its draft National Artificial Intelligence Strategy at an International Telecommunication Union (ITU) workshop in Kenya in May 2026. INTIC opened the first version to public consultation, initially setting a 4 June deadline before publishing a later consultation notice with a deadline of 4 September.

The proposed strategy identifies opportunities across healthcare, education, agriculture, finance, energy, climate resilience and digital public services. Mozambique has also highlighted regulatory sandboxes, controlled environments in which emerging technologies can be tested under defined conditions, as tools for learning how AI should be deployed safely and responsibly.

These priorities could help the country address practical development challenges. AI-supported agricultural services, for instance, could improve access to information for farmers, while carefully designed public-sector systems could help institutions process information and deliver services more efficiently. But the value of such applications will depend on the quality of the underlying data, the availability of reliable infrastructure and the capacity of public institutions to evaluate results.

The crucial question is whether the strategy will move beyond identifying promising sectors to establish a credible implementation plan. That means defining priorities, assigning responsibilities, identifying funding sources and setting measurable targets.

Without those elements, a national strategy risks becoming a statement of intent rather than an instrument of economic transformation.

As of 22 October 2026, the official material reviewed does not confirm that the final strategy has been formally adopted by the Council of Ministers. The draft therefore signals the government’s policy direction, but should not be treated as a binding, final national AI framework.

One of the CNIA’s most important functions is to help coordinate the institutions responsible for different aspects of AI governance. The commission includes representation from public authorities, the judiciary, regulatory bodies, research institutions and other stakeholders. This breadth could help prevent AI policy from developing in disconnected parts of government.

However, coordination is not the same as implementation.

AI-related questions can fall across several existing areas of regulation, including data protection, cybersecurity, financial supervision, communications and public administration. A single AI application may raise issues in several of these areas at once. Without clear institutional responsibilities, businesses may face uncertainty about which authority to consult, while public agencies may struggle to determine who is accountable when an AI-assisted decision goes wrong.

Mozambique has already advanced parts of its wider digital regulatory framework. INTIC lists the Cybersecurity Law and Cybercrime Law, both approved on 1 July 2026, alongside regulations governing data centres and cloud computing adopted in late 2025.

These instruments provide important foundations for a more secure digital economy, but they do not automatically resolve every question raised by AI. Cybersecurity rules, for example, cannot alone determine when an automated decision requires human review or how institutions should assess the fairness and reliability of an AI system.

The CNIA could help connect these separate frameworks, identify gaps and recommend proportionate rules. Its influence, however, will depend on how its advice is translated into decisions by the government and the relevant regulatory authorities.

Mozambique is also positioning itself as a potential destination for data-centre and cloud-computing investment. At the May ITU workshop, INTIC highlighted the country’s electricity-generation potential, water resources, coastline and submarine cable opportunities as advantages for digital infrastructure development. The government also invited private-sector investors to consider the country for data-centre projects.

The economic logic is understandable. AI systems depend on computing infrastructure, connectivity and access to data. Investment in these foundations could strengthen digital services, support local enterprises and improve Mozambique’s ability to participate in the wider technology economy.

But the distinction between hosting infrastructure and building an AI industry is important.

A data centre can provide computing capacity without necessarily creating a strong domestic ecosystem of AI developers, researchers, startups and skilled workers. The broader economic return depends on whether local institutions and businesses can access that infrastructure, develop useful applications and retain a meaningful share of the resulting value.

There are also trade-offs that deserve attention. Large infrastructure projects can place demands on electricity, water and land. Their contribution to national development should therefore be assessed not only through investment figures, but also through employment, skills transfer, affordability, environmental sustainability and opportunities for local suppliers.

For the CNIA, this creates a strategic question: how can Mozambique attract international technology investment while ensuring that infrastructure development supports domestic innovation and wider economic participation?

The answer requires coordination between AI policy, energy planning, education, research and industrial development, not simply incentives for technology investors.

Responsible AI adoption also requires more than technical performance. Systems used in healthcare, education, finance or public administration may affect access to services, livelihoods and personal information. Errors, biased outputs or poorly designed automated decisions can impose real costs, particularly where affected individuals have limited ability to challenge the outcome.

The CNIA’s mandate to examine ethical, legal and security issues provides an institutional starting point. The next step is to translate these principles into practical expectations for organisations developing or deploying AI.

Those expectations could address risk assessment, data governance, human oversight, transparency and mechanisms for reviewing consequential decisions. The precise obligations should emerge through the appropriate legal and policy processes rather than being assumed to exist already.

Public participation matters as well. INTIC’s consultation process offered stakeholders an opportunity to comment on the proposed strategy, and the August engagement with Vodacom Mozambique demonstrated efforts to gather private-sector input.

But meaningful consultation requires more than inviting submissions. Stakeholders need a reasonable opportunity to contribute, and the final policy should make clear how competing concerns were considered. Smaller businesses, researchers, educators, civil society organisations and communities affected by AI adoption should not be overshadowed by the interests of large investors or public institutions.

The decree’s provision for publishing the CNIA’s reports, opinions and recommendations every six months on a publicly accessible digital platform could help strengthen accountability. Its value will depend on whether the reports are published consistently and provide substantive information about the commission’s work and the government’s response.

Mozambique is developing its AI governance framework as other African countries pursue their own strategies and regulatory approaches. Regional cooperation could help it learn from different policy models, share expertise and reduce the cost of developing technical and regulatory capacity independently.

The ITU workshop in Kenya provided one platform for this exchange, bringing Mozambique together with representatives from countries including Tanzania, Zambia and Cameroon. International organisations and development partners are also supporting Mozambique’s strategy-development process.

Such cooperation can accelerate learning, but international frameworks cannot simply be transplanted without considering local conditions. Language diversity, uneven connectivity, limited access to specialised skills and differences in institutional capacity all influence how AI systems perform and who benefits from them.

Mozambique’s policy challenge is therefore twofold: align with useful international principles while ensuring that its approach reflects national priorities and practical constraints.

This is particularly important if AI is to improve services beyond major urban centres. An advanced system that depends on expensive connectivity, extensive digital records or specialist staff may have limited impact where those foundations are absent. Building basic digital capacity alongside AI adoption is not a secondary task; it is part of making the technology useful.

For businesses, the CNIA’s establishment signals that AI governance is becoming a more structured part of Mozambique’s digital policy agenda. Companies developing, integrating or deploying AI should monitor the final national strategy, subsequent legislative proposals and any guidance issued by the relevant authorities.

The immediate priority is to distinguish policy signals from binding obligations. The commission’s creation and the publication of a draft strategy do not, by themselves, impose a comprehensive set of new AI-specific compliance requirements on every business.

For investors, the question is whether Mozambique can offer both opportunity and predictability: a credible path for infrastructure development, clear regulatory responsibilities and a workforce capable of supporting technology-intensive activities.

For government, the measure of progress should be concrete outcomes. These could include funded pilot projects, improved public services, stronger research capacity, workforce development and transparent evaluation of AI systems in use. The relevant metrics will vary by sector, but they should make it possible to distinguish genuine improvements from announcements and activity alone.

Mozambique has taken an important institutional step by establishing the CNIA and advancing a draft national AI strategy. It is also developing the wider digital infrastructure and legal framework on which responsible AI adoption will depend.

But these developments represent the beginning of the process, not proof that an effective AI ecosystem is already in place.

The next milestones are the formal adoption of the strategy, a clear implementation timetable, defined institutional responsibilities, credible financing and regular public reporting on progress. The CNIA’s first recommendations will be particularly important in showing whether it can connect policy ambition to decisions that institutions and businesses can act upon.

Mozambique’s AI opportunity will not be measured by the number of strategies it publishes or the scale of infrastructure it attracts. It will be measured by whether those investments and institutions produce clearer rules, stronger local capabilities and tangible benefits for citizens.