본문으로 바로가기

NEWS
BACK TO LIST

[GIB INSIGHTS]Global Perspectives on AI Regulation: Current Trends and Future Directions for AI regulation in Nigeria

  • Date 2025-01-07 16:20
  • CategoryResearch and Education
  • Hit509

[GIB INSIGHTS]Global Perspectives on AI Regulation: Current Trends and Future Directions for AI regulation in Nigeria 사진1

Artificial intelligence is arguably one of the disruptive technologies of the 21st century. It will shape how we do things as an essential part of our daily lives (Ubena, 2022). It has been deployed in several sectors, including the military, health, business, manufacturing, and agriculture. 

Due to its intrusive nature, its regulations have been controversial, with different jurisdictions adopting several approaches. The market-driven approach to AI regulations in the United States focuses on self-regulation and voluntary reporting. This approach seeks to create an enabling environment where AI innovation can thrive without being stifled by regulations. In the United States, the National AI Initiative Act of 2020, the 2023 Executive Orders on AI and the White House Blueprint for an AI Bill of right provide a supportive environment for advancing AI development.

The European Union, on the other, adopted the rights-driven/risk-based approach by enacting specific AI legislation that seeks to ensure a balanced development of AI in accordance with human rights (Kop, 2021). This approach classifies the applicable risk criticality and provides permissible and impermissible use/deployment of AI.

The third approach is the state-based approach adopted by China, which seeks to provide state-led enablement for AI growth, provided it is used in furtherance of the state policies and objectives. Here, the state is the regulator (e.g., the 2021 regulation on recommendation algorithms, the 2022 rules for deep synthesis (synthetically generated content), investor, end-user; and policy maker ( e.g., the proposed 2023 draft rules on generative AI),  (Sheehan, 2023).

Flowing from US and EU approaches mentioned above, the underlying idea behind the regulatory approach suggests a stronger desire for data protection, provision of liability thresholds, and transparency. Arguably, these are not the defining characteristics of the Chinese approach to AI regulation, the Chinese regulatory framework is primarily characterised by state control, alignment with national policies and a focus on ensuring that AI technologies serve the interest of social stability, political objectives and economic development. 

Nigeria, as with other developing economies, does not have any specific AI legislation; the attempt at AI regulation is the Draft National AI Strategy of 2024 recently issued by the Federal Ministry of Communication, Innovation and Digital Economy (FMCIDE). This comprehensive document outlines the Nigerian Government's approach to AI and the need to utilize the immense opportunities offered by AI, its implementation and allowable trade-offs shaped by societal values, and the specific outcomes the Nigerian Government wants to achieve. The policy adoption of the U.S. National Institute of Standards & Technology (NIST) Framework for AI Risk Management is commendable.

As stated in the draft policy document, a review of 50 national strategies revealed three central archetypes (a. national enabler, b. Specialist -AI ambition aimed to develop specific expertise that can be promoted globally); and (c.) Industry Leader -AI ambition archetype that focuses on being global leaders). They are pretty similar to the three approaches explained above.

The critical difference thereof is that the draft policy seeks to be a jack of all trades and master of none. The strategic pillars of the draft policy are listed below, along with well-written/articulated actionable initiatives. 

a. Building Foundational AI Infrastructure

b. Building and Sustaining a World-class AI Ecosystem

c. Accelerating AI Adoption and Sector Transformation

d. Ensuring Responsible and Ethical AI Development

e. Developing a Robust AI Governance Framework

The above appears to balance all aspects of the AI sphere without any significant focus or trade-offs. From the models explored above, there are significant trade-offs for each approach. The American model seeks to ensure AI leadership at the expense of effective social regulation for AI harms such as bias and discrimination. The EU model ensures responsible and ethical AI development while stifling AI innovation (Judicial ban on predictive policing in Germany (Solderholm, 2023). In contrast, China’s approach focuses on AI use and development at the expense of human rights/privacy protection. 

For a developing country plagued with low data collection rate, poor infrastructure, there is an absence of high-performance computing (HPC) resources that scale local AI development and low private sector investment in AI infrastructure. The optimal or realistic approach would be to consider a hybrid approach- that combines the strength of the U.S. and EU models. Nigeria should adopt the EU model to ensure that the use or deployment of AI technologies does not violate its citizen's fundamental human rights or promote existing or potential tribal bias or discrimination. This is quite important given that most biases in AI prediction models do not originate from the algorithm but from the dataset used in training the models. 

For a country with a long history of deeply entrenched tribalism due to ethnic diversity (Agbede & Oparinde, 2024), the focus should be on the responsible adoption and ethical use of AI to accelerate development and avoid in its entirety any form of representation bias that may occur from datasets embodying the current systemic bias. 

The adoption of the U.S. model will ensure the implementation of innovation friendly policies such as public private partnership, voluntary compliance and incentives and creation of regulatory sandboxes for AI start-ups. This is important because realistically, the absence of the necessary technological infrastructure has placed us in a disadvantaged position in the AI leadership race. A recent study states that Nigeria ranks second in Africa with over four hundred (400) AI startup firms (Centre for Intellectual Property and Information Technology Law, 2023). Similarly, a recent report from Google reveals that Nigeria has the sixth-highest search interest in Artificial Intelligence. 

Even though the reports states that Nigeria a famous market for AI, we should not be comfortable by leading at the consuming side of the chart, we ought to develop our own AI ecosystem based on our own data and fine-tuned to cater for our diverse society and peculiar socio-economic problems including the protection of vulnerable groups.

This hybrid approach will enable Nigeria to foster a healthy AI ecosystem that promotes innovation while mitigating AI related harms to the AI consumers/end-users in Nigeria. Rather than dissipate energy/ available resources on unattainable goals as contained in the draft policy document, a more structured focus on AI development and safety will develop Nigeria’s AI ecosystem and ensure that we become a reference point for AI research and consumer protection.


References 

Agbede, Grace & Oparinde, Kunle. (2024). Tribalism: A Thorny Concern in Nigerian Politics-A Discursive Review and Appraisal. 18. 46-55. 10.51709/19951272/Spring2024/4. 

Centre for Intellectual Property and Information Technology Law. (2023). The state of AI in Africa Report 2023. https://cipit.org/wp-content/uploads/2023/06/Final-Report-The-State-of-AI-in-Africa-Report-2023.pdf

Kop, M. (2021). EU Artificial Intelligence Act: The European Approach to AI, Transatlantic Antitrust and IPR Developments. Stanford University, Issue No. 2/2021. https://law.stanford.edu/publications/euartificial-intelligence-act-the-european-approach-to-ai/

Sheehan, M. (2023). China’s AI Regulations and How They Get Made. Carnegie Endowment for International Peace. https://carnegie-production-assets.s3.amazonaws.com/static/files/202307-Sheehan_Chinese%20AI%20gov-1.pdf

Söderholm, S. (2023). Fundamental rights control when implementing predictive policing – a European perspective. Peking University Law Journal, 11(1), 91–104. https://doi.org/10.1080/20517483.2023.2223850

Ubena, J. (2022). Can Artificial Intelligence be Regulated? Lessons from Legislative Techniques. In L. Colonna & S. Greenstein (Eds). Nordic Yearbook of Law and Informatics 2020–2021 Law in the Era of Artificial Intelligence Liane. The Swedish Law and Informatics Research Institute. 295–314. 


Visit 2024 Global Issue Brief Insights: Policy Lecture Series Website [[CLICK HERE]]

Related News



Most Read