As the role that artificial intelligence plays in society increases, African nations must factor the technology into their national security plans, according to experts in the field.
“AI strategy can be considered a sectoral strategy within the larger security framework,” Joel Amegboh, a security expert at the African Center for Strategic Studies, said during a recent webinar. “That could allow governments to integrate AI as a critical tool to enhance national intelligence, law enforcement, national defense or economic development.”
AI has the potential to improve defensive capabilities, whether the task is predicting terrorist attacks or forecasting maintenance schedules for key equipment. The technology already has shown itself to be effective in predicting and enabling responses to pirate attacks in the Gulf of Guinea.
The African Union’s Peace and Security Council held its first meeting to discuss the impact AI could have on the continent’s peace and security in 2024. It followed that with a second discussion on March 20. One focus of both discussions was the risks associated with the rapid development if AI in an environment with few regulatory controls.
Despite the advantages AI offers, many leaders continue to hold it at arm’s length, whether due to a lack of understanding or a mistrust of the technology. For that reason, it typically needs a champion in the government to drive interest in an AI strategy and to bring political leaders together on the issue, according to Amegboh.
“It’s important to ensure the strategy is specifically tailored to the needs of the country implementing the strategy to ensure the strategy is inclusive and equitable,” he added.
The United Nations Institute for Disarmament Research is compiling a global collection of national AI policies and approaches at the website http://www.aipolicyportal.org. The U.N. has begun developing guidelines to help countries integrate AI technology into their defense plans, according to Yasmin Afina, a researcher for the security and technology program of the institute.
“The guidelines do not aim to be proscriptive,” Afina said during the Africa Center webinar. “Rather, we’re hoping the draft guidelines will be a toolbox.”
One of the crucial techniques in that toolbox is developing the local human capital to develop and operate AI systems, she said. “We need to train and retain AI talent. We also need to train staff on technical and ethical issues of using AI in defense and security contexts. In Africa, paying attention to international law and ethics is crucial.”
AI’s ethical concerns are rooted in the data used to train the technology, Afina noted. Most AI programs are developed outside Africa, raising issues with the social and cultural nature of the data underlying the technology.
“Policing data that may be highly biased could taint AI future results,” Afina said. “It’s not just about data being available. It about data being curated and being useful for whatever you are doing.”
African nations are at different stages in developing their own AI strategies. Some have chosen to make their AI components public; others are keeping them secret. Kenya recently released a draft of its national AI strategy for public comment. That approach promotes transparency about the process and considers the intersection of security and human rights, Afina said.
“It’s good to consult with members of the public to ensure the concerns of academic experts and the general public are taken into account,” Afina said.
Ultimately, a broad approach is vital to developing an effecting AI strategy, according to Afina and Amegboh. That means both getting buy-in from all levels of the government as well as working across agencies to guarantee the data underpinning the technology is valid and suited to the purpose for which it is used.
“Countries need to ensure that AI comes with collaboration across the government to address national challenges,” Amegboh said. “It is crucial for ministries and departments to coordinate so they can benefit from the AI strategy.”