Promoting Effective and Inclusive ICT Policy in Africa

Tag Archives: AI

  HomeCollaboration on International ICT Policy for East and Southern Africa (CIPESA)  /  AI
08Jul

What Global South Civil Society Wants from AI Governance

Author CIPESA Posted on

By CIPESA Writer |

As global discussions on the future of Artificial Intelligence (AI) governance take place at the AI for Good Global Summit and the Global Dialogue on AI Governance, questions about who shapes AI systems, whose interests they serve, and how affected communities can participate in decision-making are becoming increasingly urgent.

The Collaboration on International ICT Policy for East & Southern Africa (CIPESA) is pleased to share this joint statement by the Global Digital Justice Forum and the Global South Alliance, of which it is a member. The statement reflects concerns that CIPESA has consistently raised through its research and policy engagement, namely, current approaches to AI development risk deepening existing inequalities, and meaningful AI governance requires stronger corporate accountability, equitable data governance, and investment in public-interest AI infrastructure.

Through submissions to national AI strategies in Africa, analysis of AI governance trends across 14 African countries, and engagement with global AI policy discussions, CIPESA has consistently advocated for inclusive, rights-based approaches that ensure communities most affected by AI developments have a meaningful role in shaping its future.

The statement below brings together civil society perspectives from across the Global South and calls for an AI governance approach grounded in human rights, equity, public interest, and meaningful participation.

Joint Statement issued by the Global Digital Justice Forum and the Global South Alliance in the lead-up to the Global Dialogue on AI Governance

July 2026

The current trajectory of Artificial Intelligence (AI) innovation has consolidated the neocolonial structures of development. Today, a handful of US and Chinese transnational corporations dominate global AI systems. Driven by massive capital, semiconductor manufacturing dominance, and hyperscale cloud infrastructure, these companies control over 90% of global AI data center capacity. Their market capitalization exceeds the combined national income of many countries in the Global South. The wealth and power amassed by these corporations come at a staggering cost, borne disproportionately by the South. From the devalued, dehumanizing labor that is essential for training AI models to the critical minerals, land, energy, and water, communities in the South continue to provide the scaffolding for the AI economy and society, without the voice and power to shape and benefit from this paradigm. These systemic injustices also perpetuate deep dependencies on current and future infrastructures — over which communities lack control and sovereign agency.

The Global Digital Justice Forum (GDJF) and the Global South Alliance (GSA) believe that the emerging AI order lacks legitimacy; it grants unbridled impunity to powerful corporations, while reducing humanity and nature to objects of limitless extraction. The many summits and conversations about AI governance have failed to tackle these core issues. 

Against this backdrop, we exhort the UN Global Dialogue on AI Governance to deliver on a South-led AI paradigm, anchored in a vision of rights-based development, respectful of planetary boundaries, and committed to intergenerational justice and human rights. We urge that the Global Dialogue on AI Governance commit to the following.

  • End AI extractivism 

A ‘move-fast-break-things’ approach to digital innovation aids profit, not people. In particular, the systemic and collective risks and harms associated with the violation of human rights, the erosion of democratic processes, the abuse of the environment, and the discrimination and invisibility of marginalized citizens in AI-driven decision-making in public services remain consistently ignored and underplayed in international consensus declarations. AI innovation must embrace the precautionary principle. It must be ethically and transparently developed, democratically accountable, and grounded in a globally agreed minimum floor for meaningful and dignified work, pluralistic knowledge, diversified economies, and planetary flourishing.

  • Apply the Common But Differentiated Responsibilities (CBDR) principle in international AI cooperation

The reckless path of data and AI technologies, designed and controlled by a few, has led to predatory value capture, strengthening the geo-economic and geo-political power of a handful of corporate actors and countries. The human and planetary costs arising from such opportunism are indeed a common concern. However, power diff erentials in international economic law have led to a status quo where trade, taxation, and Intellectual Property regimes clearly disadvantage developing countries, disproportionately enabling a massive transfer of wealth from the South to the North. This seriously undermines the development of digital infrastructure and human and institutional capabilities in developing countries. Such asymmetry must be remedied through global commitments to underwrite the development of regenerative, locally-led, AI infrastructures and models in the South.

  • Address corporate impunity in data and AI value chains

A global moratorium on the sale and use of AI systems that pose a high risk to human rights (such as remote biometric recognition, social scoring, spyware, and AI-driven autonomous weapons) is urgently needed. The proposed UN Binding Treaty on Transnational Corporations (TNCs) to hold global businesses accountable for human rights violations and environmental degradation in supply chains needs to be adopted without delay and appropriately future-proofed against the specific risks of harms and abuses in data and AI value chains.

  • Design a data governance framework that delivers on global equity

A ‘one-size-fits-all’ policy playbook for cross-border data flows governance will not deliver on equitable development. Development sovereignty must be recognized as a core principle in the global governance of cross-border data flows. Furthermore, the governance of the non-personal data commons requires a societal approach that includes safeguards for collective privacy and the rights of communities to steward the use and re-use of their data resources in innovation ecosystems, together with strong personal data protection rights.

  • Invest in the development of global public compute

The foundational infrastructure of compute is controlled by a few corporations. Even open-source AI models are often dependent on closed/proprietary infrastructure systems for their hosting and distribution. To ensure that data science and AI innovation deliver on public innovation, a global facility for public compute is needed. A ‘CERN for AI’ could support a distributed network of AI research centers coordinated by a central hub and provide access to innovators and researchers from developing countries.

The current trajectory of AI innovation is not working for the majority. The Global Dialogue on AI Governance must move the needle with conviction and courage towards people’s participation, planetary wellbeing, and public value. Anything less will not do justice to the people of the South.

Please find the links to prior submissions from GDJF and GSA to official consultations of the Global Dialogue below:
● GDJF’s April 2026 submission
● GSA’s April 2026 submission

25Jun

Accelerating Digital Accessibility and Solutions for Africa’s Future

Author CIPESA Posted on

By Raylenne Kambua |

More than 150 million persons with disabilities across Africa navigate a digital landscape not designed for them, with inaccessible websites, unusable mobile applications, unreachable government services, and educational platforms that often lock them out. This exclusion carries a direct economic cost, which the World Bank estimates at 3–7% of countries’ Gross Domestic Product.

Increasingly, connectivity for persons with disabilities has become less a question of infrastructure coverage and more about accessibility and meaningful usage. While internet services are available to 85% of Africa’s population, 64% of those within coverage do not use them, with persons with disabilities among the most excluded groups. The exclusion has been reinforced by the high costs of devices, with some markets taxing entry-level smartphones up to 50%, making them unaffordable for low-income households.

Figures from the Assistive Technology Landscape in Africa Report show that only one in ten people who need assistive technologies across the continent have access to them, while 85% of mobility devices are still imported. This is a reflection of weak local production systems and heavy reliance on external supply chains.

At the seventh Inclusive Africa Conference, convened by inABLE in Nairobi from June 2–4, 2026, conversations examined whether Africa’s fast-expanding digital economy works for everyone or reproduces recurring forms of exclusion.

As Irene Mbari-Kirika, Executive Director of inABLE, noted, many technologies continue to fail because they are not developed with persons with disabilities in mind. This means that such devices can not be used by millions of potential users, and it has direct consequences for several users’ financial autonomy, privacy, and safety, especially in digital financial services such as mobile money.

Africa’s challenge is therefore not only access to assistive technologies, but the absence of a coherent local ecosystem for their design, production, distribution, and implementation. Building a sustainable assistive technologies value chain grounded in local materials, regional manufacturing, and culturally and linguistically relevant design is increasingly central to closing this gap.

The structural barriers to inclusion are deeply embedded across sectors, including in the education sector, where teacher training often excludes digital accessibility, curricula are rarely tested with assistive technologies, and assessment methods continue to assume uniform modes of learning and expression. These gaps, including digital literacy gaps, are compounded by the limited availability of African-language datasets, particularly for learners with communication disabilities, which constrains the development of inclusive digital and artificial intelligence (AI)-enabled learning tools.

Notably, AI presents both opportunity and risk for digital inclusion. The outcome depends on whether inclusion is embedded in design, data, and deployment. On the upside, AI-enabled tools have expanded access for blind and low-vision users, while applications in healthcare are widening access to psychosocial support on a continent with less than two mental health professionals per 100,000 people. Real-time sign language translation and voice-to-text tools are also creating new pathways for participation.

At the same time, without targeted upskilling and bias audits, AI risks simultaneously opening one door for persons with disabilities while closing others. For example, AI systems are automating roles such as data entry, transcription, and customer service, which have historically provided key employment pathways for persons with disabilities. Yet, AI models trained on biased or unrepresentative datasets and automated decision-making processes risk excluding persons with disabilities from recruitment systems.

Mercy Ndegwa, Director of Public Policy for Africa at Meta, stressed this point, noting that AI systems can only reflect communities whose data and voices are included in their design and training. This makes the inclusion of organisations of persons with disabilities in AI governance not only a rights imperative but also a technical requirement for building functional systems.

However, African-language datasets remain severely underdeveloped,  and the cost of building them at scale is prohibitive for most actors.

The launch at the summit of the development of Africa’s first Harmonised Digital Accessibility Standard for ICT Products and Services marks an important step toward continent-wide alignment. The 24-month participatory process targets adoption across all 45 African Organisation for Standardisation (ARSO) member countries. The regional standard will be adapted to African languages, culture, and infrastructural realities, changing the procurement baseline for governments while setting a compliance reference for private technology developers across the continent. Fourteen countries have reportedly confirmed participation.

Throughout the discussions, the principle “Nothing about us without us” remained central, with an acknowledgement that persons with disabilities are contributors, decision-makers, and leaders in designing systems that affect them, and not merely end-users or research subjects to be consulted only after decisions are made. Design consultant Rama Gheerawo framed this through three registers: designing for, designing by, and designing with persons with disabilities as co-creators throughout all processes.

CIPESA has been emphatic that governments, regulators, and telecommunication operators bear the greatest responsibility for digital inclusion for persons with disabilities. Civil society efforts cannot substitute for enforceable action and measurable implementation. Limited capacity to engage in technical standard-setting also continues to hinder progress on digital rights for persons with disabilities. At the same time, fragmented approaches and shifting donor priorities are placing increasing strain on the sustainability of this work.

The European Union (EU) AI Act sets a benchmark for how regulation can protect marginalised groups, while the African Union (AU) Continental AI Strategy provides guiding principles on how AI should be developed, governed, and used across the continent. However, no African has developed definitive AI legislation, although several are developing policies or strategies. As CIPESA has emphasised, AI policy frameworks across Africa must include persons with disabilities from the outset.

More specifically, efforts must be geared toward fixing AI at the source by including representatives of persons with disabilities in training data, ensuring algorithmic accountability, and making AI-powered public services accessible to all. Moreover, all AI regulations and policies must have explicit disability provisions.

Additionally, governments must reduce sector-specific taxes on entry-level assistive devices and assistive technology hardware, and mandate that Universal Service Access Fund disbursements include explicit, measurable targets for connectivity for persons with disabilities.

22May

Zimbabwe’s National AI Strategy: Policy Lessons for Africa

Author CIPESA Posted on

By Edrine Wanyama |

Zimbabwe recently adopted its National Artificial Intelligence (AI) Strategy 2026–2030 (AI strategy)  to guide digital technology and transformation in the country. The strategy aims to accelerate development, enhance industrialisation, and improve service delivery in sectors such as health, finance, agriculture, education and public administration. The strategy emphasises building local data infrastructure as opposed to relying on foreign data storage infrastructure while promoting an AI governance approach grounded in Ubuntu, human rights, accountability, transparency and inclusivity.

However, an important question is whether Zimbabwe’s approach offers useful lessons for other African countries developing national AI strategies.

Lessons for Other African Countries

The country’s AI strategy is organised around six pillars that together map a practical path for AI adoption and deployment. First, AI talent and capacity development is essential for ensuring that institutions have the skills needed to implement AI effectively. Second, AI infrastructure and computational sovereignty are necessary for ensuring digital and data sovereignty. Third, AI adoption and service transformation are critical for supporting the integration of AI across public and private sectors to improve their productivity, accountability and transparency.

The fourth pillar, AI governance, ethics and regulation, is essential for building public trust and creating a framework that supports responsible innovation. The fifth pillar, AI research, development, and innovation, can drive investments, expand knowledge production and strengthen academic output. The sixth pillar, strategic international collaboration, presents an opportunity for global partnerships with key players and stakeholders, technology exchange, and potentially greater investment.  

Consequently, these pillars offer useful lessons for other countries seeking to harness AI for socio-economic transformation while protecting data rights and data sovereignty.

Alignment with the African Union (AU) AI Strategy

Zimbabwe’s AI Strategy reflects several priorities contained in the AU Continental Artificial Intelligence Strategy, particularly the emphasis on coordinated AI governance, digital sovereignty, and sectoral innovation. Zimbabwe’s strategy aims to harmonise the deployment and use of AI across sectors such as health, finance, agriculture, education and public administration through common governance benchmarks for AI governance. If implemented effectively, these goals could help to address digital neo-colonialism, an issue that has been dominant in Africa’s technological space.

The Strategy also places strong emphasis on AI as a tool for socio-economic development, aligning with Agenda 2063 and the Sustainable Development Goals (SDGs), particularly in sectors such as health, agriculture, and education. The Strategy promotes the deployment of AI to improve agriculture through crop disease prevention, as well as mining and mineral development, which is consistent with the AU AI strategy’s priorities on resource optimisation and climate resilience.

However, Zimbabwe faces significant governance and implementation challenges. The country scored 0 in the 2024 Global Index on Responsible AI Governance, highlighting the gap between policy ambition and institutional readiness. This means it requires major actions to implement the strategy, such as the establishment of robust legal safeguards, accountability mechanisms, oversight institutions, and rights-based governance frameworks, which are also emphasised within the AU strategy.  Other African countries can draw lessons from Zimbabwe’s approach, such as the need to complement AI strategies with stronger governance capacity, clearer regulatory safeguards, and more coherent data governance frameworks to support responsible and accountable AI deployment.

UNESCO Guidance on AI

The UNESCO Recommendations on Ethics of Artificial Intelligence, adopted in 2021, is a global normative framework that promotes human rights, including human dignity, transparency, fairness, human oversight in AI systems, and democratic participation. It also provides practical policy action areas covering issues such as data governance, gender, education and research, health, and social wellbeing.

While the UNESCO Guidance is emphatic on ethical and privacy considerations, Zimbabwe’s strategy falls short. Ambitions to integrate AI into public service delivery sectors such as education, health, and public administration will require stronger safeguards to ensure alignment with the human-centric principles articulated in the UNESCO framework. In the age of AI, data security concerns, intellectual property rights, algorithmic bias, and institutional accountability are central to responsible deployment of AI and require clearer policy and regulatory attention.

Similarly, the UNESCO Guidance warns against the use of AI in a manner that undermines democratic participation, civic engagement, and collective decision-making. This is especially important in contexts where surveillance technologies such as facial recognition, drone monitoring, communication tracking, and social media surveillance are deployed without clear safeguards or independent oversight. Zimbabwe, like several other African countries, has invested in AI-enabled infrastructures, such as the “smart city” systems to monitor and surveil citizens in ways that are opaque and lack clear accountability mechanisms.

As African countries continue developing national AI strategies and governance frameworks, they must strive to ensure that the deployment of AI is transparent, publicly accountable, and pays close attention to ethical and human rights standards. Without these safeguards, AI risks reinforcing exclusion, surveillance, and digital authoritarianism rather than advancing development.

Conclusion

Zimbabwe’s adoption of an AI Strategy is an important step toward advancing tech-enabled digital and socio-economic transformation. It also reflects the country’s intent to align national priorities with the African Union’s vision for AI-driven development across the entire continent. However, for such strategies to be effective and legitimate, they must be grounded in ethical and human rights standards laid down in regional and international benchmarks.

12May

 How the WHO Digital Health Strategy Should Govern Data, AI and Digital Public Infrastructure

Author CIPESA Posted on

By Raylenne Kambua |

As the World Health Organization (WHO) develops the Global Digital Health Strategy for 2028–2033, the Collaboration on International ICT Policy for East and Southern Africa (CIPESA) has submitted recommendations urging that the strategy be anchored on human rights, equity, and accountability, alongside technological innovation. 

Across Africa, Artificial Intelligence (AI), telemedicine, disease surveillance systems, and automated diagnostic systems are transforming healthcare delivery. However, CIPESA pointed out in the submission to the WHO Regional Office for Africa that technological innovation without proper governance can worsen exclusion, undermine privacy protections, and reinforce inequalities in healthcare delivery.

The submission comes at a time when key global and regional digital health governance frameworks are being reshaped. Last year, the World Health Assembly extended the Global Strategy on Digital Health 2020–2025, and simultaneously mandated a successor framework to be completed in 2027. 

Furthermore, global initiatives such as the World Summit on the Information Society (WSIS) and the Global Digital Compact emphasise that digital transformation must integrate the Sustainable Development Goals and ensure inclusive development.

At the continental level, the Africa Centres for Disease Control and Prevention (Africa CDC) has rolled out the Africa CDC Digital Transformation Strategy which alongside the  African Union (AU) Data Policy Framework advances interoperability, transparency, privacy, and the ethical deployment of digital systems in the health sector. However, CIPESA notes that despite these commitments, implementation gaps remain significant, particularly regarding health data governance, accountability, and protection against algorithmic harm.

CIPESA’s work on health data governance in Uganda, patient data privacy in Ghana, Rwanda, and Uganda, and analysis of Kenya’s Digital Health Act, point to the same reality. The rules governing who controls health data, who is included in digital health systems, and who is held accountable when data is mishandled are still weak across most of the continent.

“As countries embrace AI, digital public infrastructure, and data-driven healthcare systems, the real test will be whether these technologies strengthen confidence in public health systems or deepen concerns about exclusion, surveillance, and the misuse of personal data,” said CIPESA Executive Director Dr. Wairagala Wakabi.

He added: “Trustworthy digital health systems require transparent digital infrastructure, accountable AI systems, and strong data protection safeguards. Africa has the chance to shape a digital health governance model that is innovative, inclusive, and based on the public interest and human dignity.”

CIPESA’s Core Positions and Recommendations

Digital health offers significant potential to enhance Universal Health Coverage and strengthen health systems across Africa. However, without governance anchored in rights, equity, inclusion, and accountability, this promise will remain unfulfilled. It is against this backdrop that CIPESA submitted the following recommendations:

1. Digital Public Infrastructure (DPI)

Digital health infrastructure should be open, interoperable, transparent, and rights-respecting, with safeguards to prevent exclusion and misuse of shared systems.

    2. Health Data Governance

    Most African countries lack specific laws that govern health data. Countries should therefore establish clear legal frameworks governing health data, including informed and meaningful patient consent, limits on data sharing, independent oversight mechanisms, and enforceable accountability structures.

    3. Artificial Intelligence (AI)

    CIPESA warns that most AI systems used in healthcare are trained on non-African datasets, which increases the risk of inaccurate diagnoses and exclusion. The submission recommends that AI tools and systems should be tested and validated in Africa, and include mandatory “explainability” standards so that health professionals understand how the AI reaches conclusions, and safeguards against bias in clinical decision-making tools.

    4. Interoperability

    Many digital health tools are isolated across countries and institutions, meaning they can not share data with each other. In this light, CIPESA recommends the adoption of national interoperability standards, including the WHO SMART Guidelines, to ensure secure and efficient health data exchange. Also, all digital health vendors should adhere to interoperability standards and the utilisation of shared infrastructure.

    5. Equity and Inclusion

    The digital divide continues to expand in most African countries and limits access to digital health services. CIPESA recommends conducting “equity impact assessments” before launching new systems, continued availability of offline options, and supporting digital literacy initiatives.

    6. Stronger Governance

    CIPESA holds that technology fails without clear leadership and coordination between health and technology departments. Therefore, creating clear governance structures for accountability and embracing a multi-stakeholder approach in decision-making processes are vital for resilient health systems. Other recommendations are the publication of institutional AI and digital health use policies and mandatory human rights impact assessments for high-risk systems.

    7. Sustainable Financing

    Many digital health initiatives rely on short-term donor funding, resulting in countries being dependent and unable to scale such programmes. Additionally, gaps in workforce capacity constrain implementation. CIPESA urges governments to invest in domestic financing of digital health systems and training of health and technical personnel.

    In conclusion, CIPESA’s submission emphasises that while digital technologies offer significant opportunities to strengthen health systems and improve service delivery, without strong safeguards, digital health risks reproducing and scaling existing inequalities in new, technologically mediated forms. A rights-based, inclusive, and accountable approach is therefore essential to ensure that Africa’s digital health future is not only innovative, but also equitable and just.

    Read the full submission here.

    11May

    Outpaced by Its Own Ambition: Can Kenya Bridge Its AI Regulation  Gap?

    Author CIPESA Posted on

    By Raylenne Kambua |

    The raw paradox at the heart of Kenya’s Artificial Intelligence (AI) moment is that the country is simultaneously sprinting ahead in AI adoption while grappling with a shrinking space for the very digital voices that AI empowers.

    According to the Digital Global Update Report, Kenya recorded the world’s highest usage rate of AI tools in 2025, with 42.1% of internet users aged 16 and above reporting active use of AI-powered technologies. This level of usage indicates that AI is increasingly being woven into the daily life of Kenyans.

    However, the Navigating the Implications of AI on Digital Democracy in Kenya report by the Collaboration on International ICT Policy for East and Southern Africa (CIPESA) highlights that while AI empowers citizens, it also enables unprecedented surveillance and manipulation.

    A Nation Leading the Way in AI Adoption

    Kenya has made significant investments in digital services, innovation hubs, and connectivity under the National Digital Master Plan 2022–2032.

    These developments are also transforming how citizens interact with the government. Tools such as the Office of the Data Protection Commissioner’s Linda Data chatbot and platforms such as Sauti ya Bajeti have expanded access to rights information and budget tracking.

    Yet, even as AI delivered clear benefits, it also revealed its dual nature, most visibly during the 2024 #RejectFinanceBill protests, during which Gen Zs mobilised through AI-generated infographics, satire, and short-form videos. At the height of the protests on June 25, a nationwide internet disruption was enforced despite assurances from the Communications Authority. The disruption was confirmed by network monitors like Cloudflare and NetBlocks, exposing the fragility of internet freedom in Kenya.

    Civil society condemned the internet shutdown as a violation of rights, while telecoms Safaricom and Airtel attributed it to outages in their undersea cable. In the aftermath, reports of abductions and enforced disappearances of digital activists escalated, with the Kenya National Commission on Human Rights documenting at least 82 cases between June and December 2024.

    Kenya’s AI Policy Landscape

    The launch of the Kenya National AI Strategy 2025–2030 in March 2025 signalled the country’s ambition to position itself as Africa’s leading AI innovation hub. The strategy prioritises governance, ethics, investment, digital infrastructure, data ecosystem development, and support for AI research and innovation.

    Kenya has also strengthened its international profile through participation in programmes such as the United Nations High-Level Advisory Board on AI, joining the International Network of AI Safety Institutes, and assuming leadership in the World Summit on the Information Society (WSIS+20).

    At the national level, initiatives such as Digital Platforms Kenya (DigiKen) and the Kenya Bureau of Standards’ draft AI Code of Practice reflect growing momentum toward operationalising AI governance and skills building. The government is also developing an AI and Emerging Technologies Policy and a Data Governance Policy, both of which are expected to be in place by July 2026.

    However, the gap between ambition and readiness remains wide. Kenya ranks 93rd in the 2025 Government AI Readiness Index, due to persistent weaknesses in infrastructure, implementation, and institutional capacity.

    Moreover, Kenya’s legal framework for AI remains fragmented and incomplete. Currently, there is no standalone AI law in force, but a controversial Artificial Intelligence Bill, 2026, that has raised significant concerns about over-regulation and censorship  is under discussion. Additionally regulation is based on broader laws such as the Data Protection Act 2019 and the Computer Misuse and Cybercrimes Act 2018, which were not designed to address AI-specific risks such as deepfakes, automated decision-making, algorithmic discrimination, or synthetic disinformation.

    As highlighted in the CIPESA report, critical gaps remain in the use of AI. These include the absence of mandatory algorithmic impact assessments, weak safeguards against AI-driven surveillance such as facial recognition, and scant measures to address AI-generated electoral misinformation. Furthermore, regulatory authorities lack sufficient capabilities to audit and monitor sophisticated AI systems, and there are no clear licensing or accountability frameworks for AI creators and deployers.

    “Without deliberate, inclusive, and rights-centred governance, AI risks entrenching authoritarianism and exacerbating inequalities.” (Navigating the Implications of AI on Digital Democracy in Kenya, 2025)

    The Way Ahead: AI Governance Focused on Human Rights

    The CIPESA report outlines a human rights–centred approach to AI governance that is built on the following key principles:

    1. Life-Centred and Human-Centred Design and Accountability: AI should support and not replace human judgment, with strong oversight to ensure transparency and accountability.
    2. Equity and Fairness: Design AI to prevent bias and expand inclusive access, especially for underrepresented groups.
    3. Transparency and Trust: Ensure AI systems are explainable, well-documented, and open to public scrutiny and challenge.
    4. Safety, Security and Resilience: Build resilient systems with ongoing risk assessments and strong protections against misuse.
    5. International Collaboration and Ethical AI Development: Advance ethical AI through international collaboration while upholding constitutional values and human oversight.
    6. Environmental sustainability: Align AI development with climate resilience and sustainable resource use.
    7. Inclusive Participation and Cultural Relevance: Reflect local diversity and involve marginalised communities in AI design.
    8. Robust Governance and Adaptive Regulation: Maintain flexible, responsive regulation that keeps pace with technological change.

    The report calls for a coordinated, multi-stakeholder approach to AI governance. It recommends that:

    • The government should enact a comprehensive AI law aligned with constitutional and international human rights standards, establish a legally mandated National AI Advisory Council with inclusive representation and strong enforcement powers.  It should also introduce clear prohibitions on high-risk practices such as real-time biometric surveillance without judicial oversight.
    • Civil society and the media should strengthen public awareness, promote accountability, and counter AI-driven disinformation.
    • Private sector actors should uphold transparency, fairness, and ethical standards across AI systems, including fair labour practices. Labour protections must be guaranteed for gig workers and data annotators within the AI value chain.
    • Academia and research institutions should continue generating evidence that can guide context-specific policy and regulation.
    • Across all stakeholders, digital literacy must be expanded, especially in underserved and rural communities, so that citizens can understand and challenge AI systems that affect them.

    With the ongoing legislative processes on AI, this is a pivotal time for Kenya, as it has the momentum and the attention of the world. But momentum without action will not work. The country cannot afford slow, fragmented debates while technology is fast progressing. Additionally, Kenya must strike a careful balance between regulation and innovation, as overly restrictive rules could limit access, slow local innovation, and lock the country out of AI’s economic and social benefits. The goal should be a flexible, forward-looking framework that protects rights while still enabling growth and opportunity.

    Read the full report, Navigating the Implications of AI on Digital Democracy in Kenya.

    1 2 3 4