Harmonizing the Techno-Verse: Unravelling AI Regulation in Kenya and the Ideal Governing Hand

Image by iuriimotov on Freepik

Introduction

Recent months have sparked a flurry of excitement and worry in Kenya's tech sector, capturing the attention of both techies and tech enthusiasts. The stratospheric growth of Worldcoin, which promised to usher in a new era but was instead halted due to data protection concerns and apprehensions about the trading of personal information, demonstrated a stunning spectacle on this tech-driven stage. Meanwhile, the halls of Parliament echoed with a symphony of opinions as a petition to ban TikTok unfolded, shining a light on the tremendous interconnectedness of artificial intelligence in our lives. Amidst the backdrop of data privacy concerns and the imperative of upholding societal values, these two scenarios have given rise to a pivotal query: the regulation of technology and, in particular, Artificial Intelligence (AI)

AI weaves intricately through human development, crossing borders and touching a wide range of domains. It frees us from manual labour, automates finance, improves healthcare, anticipates legal outcomes, and helps to design a more efficient world. One name sticks out in this symphony of technological advancements: Elon Musk. Recently, he addressed a heartfelt letter to the US government, expressing his concern about the rapid development of artificial intelligence systems. Musk sought a temporary halt in the development of new AI systems in his letter. He urged the state and stakeholders to examine the tremendous responsibilities that come with these great technologies, and instead of rushing ahead, he proposed taking a more measured approach. Musk felt that it was critical to work on developing a robust set of rules and laws for AI so that it could integrate smoothly into society. He reminded us that by properly coordinating AI development, we can unlock its enormous potential and employ it for the benefit of all.

During the first UN Security Council meeting on artificial intelligence, speakers, including the UN Secretary-General, shared Elon Musk's fears and underlined the significance of a global AI regulatory regime. They recognized AI's potential benefits as well as its concerns, including its ability to imperil international peace through the creation and unjustified use of autonomous weapons. Furthermore, most speakers pushed for national-level laws, urging governments to hold discussions and consultations to effectively oversee AI usage and ensure its appropriate incorporation into society.

The problem of Regulating AI

Owing to its special qualities, regulating AI might necessitate the creation of a new legal regime. AI systems can complete tasks in a complicated environment without constant human supervision due to their autonomous nature. This suggests that the traditional actus reus and mental state requirements for criminal offences may not necessarily apply to AI systems. Autonomous AI systems complicate predictability even further since they may respond in ways that their human counterparts did not anticipate.

Regulation issues may also arise from AI research and development techniques. Compared to other industries, AI research and development may be carried out with somewhat limited infrastructure. This is demonstrated by the large number of AI systems that have been developed in university and college residence halls, etc. It may also be challenging to determine culpability in cases where the AI system is flawed because distinct parts of the system may have been developed in various locations and jurisdictions. The AI system's internal workings might also be covert and resistant to reverse engineering. This makes it challenging for conventional tort law rules to apply to AI because of its opaque nature, also described as the "black box" characteristic of AI.

            Image by John Karsten Moran

                                                         State regulation vs. self-regulation

State regulation of artificial intelligence is vital for fostering responsible growth and innovation in AI and safeguarding the welfare of society. While self-regulation may appear enticing due to its prospects for quick decision-making in addition to expertise, it is laden with risk. The rapid growth of AI technology may exceed our ability to predict its potential repercussions, resulting in unexpected hazards and unexpected negative consequences for individuals and societies. Furthermore, self-regulation may result in a race to the bottom in which corporations prioritize profit over community benefit, possibly at the expense of safety and ethical considerations.

Government regulation, on the other hand, adopts a more balanced strategy to defend the needs of both innovators and the population as a whole. Regulatory agencies may provide an equitable playing field for responsible AI innovation by defining clear norms and standards and enabling competition that prioritizes safety and ethics. Furthermore, regulatory oversight could ensure that AI systems are created in a transparent and accountable manner, allowing the community to gain an understanding of how these technologies work and make informed choices about their adoption.

Regulating AI: Which Arm takes the lead?

Despite the pressing demand for AI systems, it's intriguing to note that Sub-Saharan Africa faces one of the lowest levels of global readiness in the realm of AI. This aligns with findings from the 2022 AI Readiness Index, a tool designed to assess governmental progress in AI implementation with a focus on three pivotal sectors: public administration, the technology industry, and data and infrastructure. Among the limited number of African nations listed in the top 100 countries according to this index, Kenya stands in the 90th position. The United Nations Conference on Trade and Development (UNCTAD) asserts that countries categorized as Least Developed Countries, like Kenya, are ill-equipped to absorb and adapt to digital transformation due to their scarce resources, less advanced technological landscape, and less productive industries. This deficiency could potentially hinder the attainment of the Sustainable Development Goals (SDGs).

Kenya has made significant progress in exploiting the potential of AI. To fully realize AI's potential, the government established a Blockchain and Artificial Intelligence Taskforce in 2018.Among the task force's suggestions was the creation of farable AI legislation while also ensuring the protection of human rights. Despite these efforts, Kenya has yet to fulfil this goal in its entirety, as current regulations, such as the Data Protection Act of 2019, the Data Protection Regulations, and the Computer Misuse and Cybercrimes Act of 2018, focus primarily on data privacy and cyber offences. Furthermore, the task force urged the establishment of an AI-friendly environment, a thorough assessment of AI-related hazards, and the implementation of risk-mitigation strategies.

The Legislature

According to the 2010 Constitution, the Legislature has the authority to make and amend laws (Article 93). This makes it an excellent tool for actively intervening in AI regulation through social and economic controls. Parliament can proactively develop policies concerning AI even before issues arise. This skill enables legislators to stay ahead of the rapidly expanding AI system, ensuring that rules keep up with advances in technology. By enacting proactive laws, the government may anticipate AI-related difficulties and opportunities and develop a complete regulatory framework.

Legislation passed by parliament reflects the will of the public since it is a democratic institution whose members are chosen by the people. In the context of regulating AI, the concept of laws passed by parliament expressing the will of the people is critical. Parliament is thus uniquely positioned to represent society's many interests and concerns while developing AI regulations. This democratic legitimacy ensures that AI-related laws are produced with the approval and input of the governed rather than imposed arbitrarily. By engaging citizens in legislation, parliament can handle the complex issues and ethical concerns linked with AI in a manner consistent with the public's beliefs and ambitions.

The potential lack of technical understanding among parliamentarians is a serious barrier to the AI legislative process. Considering AI-related issues are inherently complicated and rapidly growing, it is critical to have professionals who understand the complexities of the technology and its regulatory consequences. Kenya’s parliament, however, is not made up of technocrats. It is possible to overcome this shortcoming, however, by forming select committees consisting of AI experts. Parliamentarians can gain significant insights, data-driven perspectives, and educated suggestions by involving these professionals, allowing them to make well-informed judgments while creating AI rules. This partnership closes the knowledge gap and guarantees that legislative efforts are better prepared to face the intricacies of AI, resulting in more effective and comprehensive legislation that can protect society's interests.

The Public Petitions Committee, a special committee within Parliament, can play a pivotal role in expediting the regulatory processes surrounding AI. Given the potential hindrances posed by procedural technicalities within Parliament as well as Parliament’s lack of technocrats, the establishment of a special committee to handle such matters is critical. Citizens have already actively engaged the committee regarding AI-related discussions. In a recent petition filed with the Public Petitions Committee, content moderators raised concerns about the practices of AI developer OpenAI and its local partner, Samasource. These concerns revolved around issues of exploitation and underpayment among the outsourced employees, prompting a call for governmental investigation and regulation of the tech companies involved. The Public Petitions Committees are mandated to address these petitions within three months from receipt, underscoring their commitment to ensuring swift actions in matters about AI.

The Executive

The Executive, on the other hand, is primarily comprised of technocrats. This branch consists of departments, ministries, or organizations with a focus on specific fields, such as science and technology. These organizations have subject-area specialists who can provide in-depth insights into the technical, ethical, and sociological components of artificial intelligence. They may conduct in-depth analyses of AI trends, threats, and possibilities, ensuring that the law is up-to-date and responsive to technological advances. A dedicated AI regulatory agency, for example, can work with experts from academia, industry, and civil society to design AI laws that balance innovation with protection. Such knowledge can assist in ensuring that regulations are current, effective, and capable of handling developing difficulties. In Kenya, the Ministry of Information, Communication, and The Digital Economy, for example, may be suitable to make policies on the emerging issues of AI. The ministry has, however, not communicated any initiatives relating to or regarding AI since setting up and receiving a report from the AI and Blockchain Taskforce in 2018.

This inherent flexibility empowers the executive to swiftly propagate policies, a crucial attribute given the rapidly evolving nature of the AI industry. The expeditious implementation of regulations is of paramount importance in addressing the multifaceted dynamics of AI. Consequently, the executive branch emerges as the pivotal entity for effectively regulating AI due to its agility and capacity to respond promptly to the dynamic challenges posed by this burgeoning field.

The Judiciary

 Image on Freepik

The Judiciary, as the third arm of government, is mandated to interpret laws. According to Article 159 of the Constitution, judicial authority is derived from the people and shall be exercised by courts and tribunals established under the Constitution. The Judiciary is an instrumental tool in regulating AI in that it provides recommendations that the other two governments should consider. The judiciary, however, may not be a suitable tool for the regulation of AI in two aspects. Primarily, courts are unable to act suo moto. This means that a case must be lodged before the court for the court to act on it. This effectively nullifies the anticipatory aspect of the AI legislation.

The judiciary has recently been able to reduce case backlogs by 19%. Despite the clear progress, it is important to emphasize that due to the backlog of cases, AI-related issues may not be resolved promptly, hurting innovation. The backlog of AI cases can stifle innovation by increasing uncertainty, diverting resources, and impeding collaboration. Due to the possibility of prospective legal challenges and financial ramifications, the legal uncertainty surrounding AI-related conflicts can prevent corporations from investing in AI projects. Protracted legal disputes consume resources that could otherwise be committed to research and development, reducing the rate of innovation. Furthermore, the fear of legal challenges may deter partnerships between AI developers and enterprises, impeding cross-industry collaboration.

Additionally, while adjudicating AI-related matters, courts, including Kenyan courts, tend to place a lot of consideration on the risk of technology as compared to its benefits. This could very well hinder innovation. Courts may promote an atmosphere of caution and risk aversion among AI inventors and corporations by stressing the potential hazards and downplaying the advantages. Due to the potential legal ramifications and liabilities, this cautious approach may lead to a reluctance to invest in novel AI solutions. As a result, businesses may emphasize safe and well-established AI applications, inhibiting the development of potentially transformative AI technology. To nurture innovation responsibly, courts must adopt a balanced approach that takes into account both dangers and advantages while supporting developments in AI that can have a positive influence on society.

The unique nature of AI also contributes significantly to the difficulty of regulating it by the courts. Its black-box nature makes it particularly difficult for courts to look into a defect's cause or liability. Additionally, where an AI system created by multiple actors in different jurisdictions has a defect, the court may be unable to adequately assign responsibility. Nonetheless, due to the experience of courts in dealing with some of these issues about other forms of technology, they remain best equipped to deal with single-claim AI issues.

Way forward...

Ultimately, the Kenyan situation begs for a collective regulation of AI by the three arms of government. At the initial stage, an AI-specific Act should be drafted by Parliament. Parliament should make use of its special committees comprised of experts to advise them on the matter. The statute would then create an agency whose members would comprise skilled experts in the field of AI. The agency would be critical in formulating and implementing policies relating to the use and liability of AI in various sectors within the country. The reactive nature of the court would be best utilized in adjudicating individual court claims since these claims greatly involve the risk that accompanies such technology and do not in themselves stagger the innovation of AI.

This piece begs for a proactive stance on the part of the government to regulate AI. The government should consult and engage experts in the field and organize training initiatives to better comprehend the far-reaching implications of AI.

Natasha Kahungi is a fourth year student at the University of Nairobi Faculty of Law. She currently serves as the Editor-in-Chief of the University of Nairobi Law Journal and holds the role of Bureau Chief for Africa at Jurist. Natasha's passion lies in exploring the intersection of law and technology, which is reflected in her avid reading and writing on the subject.