The potential for AI in Africa is already being realised through several use cases. But how careful does the continent need to tread with its development?
Today, Africa represents just 2.5% of the global artificial intelligence (AI) market, but emerging applications could boost the continent’s economic growth by US$2.9 trillion by 2030, according to Artificial Intelligence for Development (AI4D) Africa.
A GSMA report titled ‘AI for Africa: Use cases delivering impact’, developed from existing research and interviews with leaders across civil society, NGOs, academia and the private sector, identifies over 90 AI use case applications in front-runner technology markets Kenya, Nigeria and South Africa that can drive socioeconomic and climate impact.
Subscribe today for free
“To harness the transformative potential of AI across Africa, there needs to be a strong focus on increasing skills for both AI builders and users, especially among underserved populations,” says Max Cuvellier Giacomelli, head of mobile for development at the GSMA.
“Better training programmes are essential, particularly in the face of a global brain drain on AI talent. To ensure Africa doesn’t get left behind, strong partnerships are required across a broad ecosystem of partners including ‘big tech’, NGOs, governments and mobile operators.”
Giacomelli adds that policies must evolve to address inequality, ethics and human rights concerns in AI deployment.
As African countries shape their own unique AI strategies, he believes that active engagement in global forums will be pivotal in defining regulatory frameworks that promote ethical AI development and safeguard societal interests, moving towards sustainable solutions that benefit all African communities.
THE USE CASES
Journalist and author Arthur Goldstuck has been writing extensively about AI over the last decade.
So when the generative AI revolution hit with the launch of ChatGPT, Goldstuck wrote and published ‘The Hitchhiker’s Guide to AI’, which outlines several AI use cases throughout the African continent.
He mentions Senegal as an early adopter of mobile apps allowing farmers to test their produce against market prices. He says a similar example has already been implemented in Kenya through a cloud-based platform called Intellipro, which allows farmers to input all data about their produce.
This is then compared to market prices and gives farmers access to loans to buy supplies like fertiliser and other essential needs.
Today, the majority of African AI use cases are related to agriculture (49%), climate action (26%) and energy (24%).
Agriculture employs 52% of the African working population and contributes 17% on average to GDP.
In sub-Saharan Africa, up to 80% of food is produced by smallholder farmers, who often use traditional techniques and lack access to information that would help improve yields.
The GSMA finds that the majority of use cases in agriculture involve machine-learning-enabled digital advisory services, which equip farmers with data-driven advice to adopt climate-smart farming practices and optimise productivity.
These solutions typically reach farmers via mobile devices, highlighting the importance of device ownership, digital skills literacy and user-friendliness.
AFRICAN AI REGULATION
But while Africa can boast several AI use cases, how carefully does it need to tread with regard to regulation of the technology?
In March, the world’s first major set of regulatory guidelines to govern AI was approved by the European Union.
“I've seen a paper produced by the South African government pointing towards the beginnings of developing an AI regulatory body, but it was very premature and very poorly informed," - Johan Steyn, human-centered AI advocate
The EU AI Act divides the technology into four categories of risk ranging from ‘unacceptable’ to high, medium and low risk.
Goldstuck says that the EU has acted as a pioneer for the regulation of AI with the introduction of the act, and for that reason it will now be treated as the model for local regulations in other countries.
Johan Steyn, human-centered AI advocate agrees, echoing the sentiment that the EU AI Act will now serve as the “gold standard for AI regulation globally”.
He notes, however, that while South Africa is a prominent part of the African continent, it is some way away from being a leading voice when it comes to AI regulation.
“Countries like Mauritius, Nigeria and Kenya have done remarkable work,” Steyn says. “I think South Africa is actually quite far behind from a regulatory point of view.”
Goldstuck goes a step further with his criticism of South Africa’s stance on AI regulation.
“I've seen a paper produced by the South African government pointing towards the beginnings of developing an AI regulatory body, but it was very premature and very poorly informed,” Goldstuck says.
In Goldstuck’s view, this was a mistake. He believes that South Africa should have waited to release something “more sophisticated and mature”, and this is something he expects to happen given the criticism of its initial framework. As a result, AI remains largely unregulated in South Africa.
Goldstuck references examples from the popular Netflix TV show Black Mirror and the Tom Cruise-led film Minority Report as examples of AI use being ‘unacceptable’. Those examples refer to predictive policing, whereby police are able to arrest people based on crimes they are likely to commit.
“That is clearly a dystopian vision of the future and I think the EU regulators must have watched a lot of science fiction because that's one of the things that they've banned,” he says.
These are elements that will be considered for local regulation, but Goldstuck says it also depends on the government.
“If it’s an autocratic government, they might like predictive policing, so we won’t see one-size-fits-all across the continent [of Africa], but we will see the EU’s AI Act serving as a model for many of these regulations.”
Yet while a one-size-fits-all act would be difficult to implement across the continent, does Africa need a union-like regulatory framework?
“I think we do,” Steyn says.
The difference, though, between the African Union and the EU is the pre-existing relationships between countries in the EU.
“I would like to have a framework for Africa, but I think it’s not something that can be dictated to all countries because there are so many diverse economies and socioeconomic challenges,” he says. “I don't think it should be a dictatorial system that everyone should abide by.
“The digital sovereignty of every state is important. Let's not recreate the wheel. Letʼs learn what we can from the EU.”
THE FUTURE OF AI
According to Steyn, there are two potential futures for AI in Africa, mirroring those in the rest of the world.
Scenario one is a potential utopian future with significant jumps in healthcare, education and economic equality. That would be a great outcome. However, the second scenario of a dystopian future is more probable in Steyn’s eyes.
“That is a future where there is total control and mass surveillance,” he says.
Steyn points to the Chinese social obedience system, which constructs a moral ranking system that will monitor the behaviour of its population and rank them based on their ‘social credit’.
The rankings are determined by China's economic planning team, the National Development and Reform Commission, the People's Bank of China and the Chinese court system, according to the South China Morning Post.
“Say the wrong things on social media, then your access to things like credit and education are minimised, and I think this could become the rule across the world,” Steyn says.
“The primary goal of any government, even in a democracy, is to stay in power – and AI is a power that previous regimes could only imagine would have ever existed.”
Steyn points to George Orwell’s 1984 as an example of what could go wrong should we give governments too much control over AI.
1984 is a cautionary tale that aimed to demonstrate what could happen if citizens allowed their governments to hold as much power as they desired.
Steyn notes that the novel was ahead of its time and can now be held as a “prophetic tale” from which we as a society can draw parallels.
Thus, he believes the dystopian view should not be considered scaremongering, given it is a future that is entirely plausible.
This is something that Steyn believes must be taken into account when considering the future of AI, not just within Africa but across the planet.
“We can’t be asleep at the wheel; we need to get it right and we don’t have 10 or 20 years to do it,” he says.
While he admits there is a lot to be excited about surrounding AI and its rapid development, Steyn says he has more to worry about for the future of the technology considering that it is still a fairly unknown entity.
He adds that AI is a foreign technology and questions if it will truly be used for a greater purpose, like closing the wider digital and economic divides that exist within societies today.
“Are we relinquishing our sovereignty, our digital sovereignty and our future to foreign powers? Or are we taking control of it?”
At this point in time, Steyn refrains from answering that question just yet.
RELATED STORIES
