According to the Rwanda 2023 AI Readiness and Maturity Framework by the Ministry of ICT and Innovation (MINICT), Rwanda is committed to developing responsible artificial intelligence (AI) principles and practices in relevant international and regional platforms. Actively contributing to operationalization.
Although AI creates many opportunities and great potential to accelerate development, many of its applications come with significant risks.
Therefore, national AI policies also include ethical considerations to capture economic development opportunities and reduce risks.
Also read: Inside Rwanda's priority areas as new AI policy takes shape
Victor Muvunyi, Senior Technologist for Emerging Technologies at MINICT, provides deep insight into the nature of AI, highlighting its decision-making abilities that are influenced by social structures.
He explained this with an example, saying, “In lending systems, AI could use data to favor individuals with stable employment, inadvertently marginalizing those with precarious employment.'' “There is,” he said. He highlighted how AI translation can introduce bias and individuals within the system may be unfairly denied a loan.
Mbunyi also pointed out that AI has the ability to generate highly convincing images, allowing for impersonation and fabrication of acts that can be falsely attributed to individuals who did not commit the wrongdoing.
He said, “To employ AI responsibly, we rely on ethical guidelines for AI, including beneficence, non-maliciousness, autonomy, justice, and explainability. These principles serve as a compass. It works, guides our actions, and allows us to carefully audit our models.” ”
Also read: UNESCO's 10 Ethics for the Use of Artificial Intelligence (AI)
To ensure the responsible use of AI, he said, “users need to understand that these AI tools rely on algorithms and data provided by companies and governments, which often contain sensitive information. Maintaining privacy is paramount, and we need to be careful when incorporating AI into our work.” process. ”
He also emphasized the need for vigilance when using AI tools, noting that users often accept their output without question as the absolute truth. He gave an example where doctors could rely solely on AI services to treat patients, which could lead to problems.
He urges users to avoid blind reliance and always cross-check the information provided by AI tools before making important decisions.
Theoneste Mlangira, an assistant lecturer in the Department of Computer Science at the University of Rwanda, said AI, like any technology, could have an impact, for better or worse. He emphasized that it has a serious impact on many aspects of life, including privacy, and that conscientious implementation is essential to ensure a positive outcome and protect individual rights. emphasized.
“In the wrong hands, AI can become a weapon and cause harm, including through cybersecurity breaches and the proliferation of deepfakes. These technologies not only destroy privacy and dignity, they also threaten discrimination. “Vigilance in the deployment of AI is paramount to protecting our society's collective well-being,” he added.
Also read: Deepfake AI content and how to detect it
As Mvunyi pointed out, Mlangira highlighted the power of AI to depict certain groups and ignore others. This selective representation highlights the importance of diversity and inclusion in AI development to ensure fair outcomes for all.
He advocates for the responsible use of AI by highlighting the important role of ethical compliance and combating bias. He emphasized the shared responsibility of data users in mitigating risks and stressed the need for collective efforts to minimize potential harm and promote a more just digital environment. did.
Odas Nyonkuru, CEO of Digital Umuganda, a Rwanda-based AI and open data company working to strengthen access to information and services in local languages in Africa, prioritizes inclusivity in his approach. Masu. He emphasized that the first step for the AI company is to curate its dataset to ensure that it properly represents both women and men, as well as individuals with different accents.
He emphasized the importance of adapting to local conditions, rather than simply rolling out solutions without considering the specific requirements of the community. He further emphasized that their efforts ensure inclusivity by focusing on local languages as well as foreign languages to prevent discrimination.
Nyoncle highlighted its company's efforts to promote fair use of AI, saying, “Our work revolves around ensuring that our clients are responsible users and addressing bias. “We have a duty to our clients to avoid expressions that favor one segment of the population over another and to ensure appropriate use.” ”
Additionally, during data collection, especially during quality assurance, we prioritize identifying and flagging bias to ensure fairness and accuracy.
