AI regulation: Why Africa must have a voice in global talks

Over the past few months it has rapidly become impossible to visit a news site without seeing stories about Artificial Intelligence (AI). According to these stories AI is reaching a tipping point and is about to revolutionise and change the way we live our lives forever.

Africans may already see AI in many aspects of how businesses and NGOs currently work, from interacting with chatbots when booking flights, to assisting shipping forecasts or using ChatGPT in our studies. But the next generation of AI programmes is predicted to be existentially different – with impacts akin to “before and after the internet” or even “before and after the use of the internal combustion engine”.

There is a degree to which I am sceptical of this hype, but at the same time there has already been a considerable number of articles on how Africa is lagging behind the rest of the world when it comes to AI development, and fears that the region will be left behind. For example, earlier this year African Business published an excellent article focusing on ChatGPT, titled “ChatGPT and the future of African AI”.

It argued that with “limited training data matching African cultural and economic realities, the output of ChatGPT could be skewed toward reinforcing Western cultural and ideological hegemony.”

The fear of European, American and Asian dominance in AI in general and programs like ChatGPT in particular is well-founded and real. Just consider the distribution of AI firms.

In 2021, in the midst of Covid-19, it was estimated that the UK had 2,000 firms specialising in AI. That figure will have increased rapidly since. The US now reportedly boasts over 13,000 AI start-ups, while China has well over 2,000. The US and China also host the world’s leading medium and large AI companies.

Jump to our continent, where a 2022 report suggested that there are over 2,400 African AI firms – with Egypt, South Africa, Nigeria, and Kenya being the major hubs. The dominance of these hubs is no coincidence. These countries have published national AI or digital strategies, and in some cases even have national AI centres and training institutes.

That aside, these sorts of comparisons are not unusual and again, not a coincidence. Despite the best efforts of many African governments, due to a combination of financial constraints and risk perceptions, the region lags the rest of the world in research and development (R&D) volumes and foreign direct investment (FDI), as well as digital infrastructure. It is therefore no surprise that the scale of AI use and development is relatively small, even if fast growing and highly innovative where it does exist.

It’s all in the regulation

But there is one aspect of AI in particular that is being left out in far too many discussions about the issue when it comes to Africa: the regulation of AI.

How the new generation of AI programs is regulated will dictate what it means to businesses and NGOs and ultimately the continent of Africa. Like so much of new technology these regulations are being decided elsewhere in the world with little consideration for the African continent.

Let me explain with one simple example.

In the 1990s there were similar predictions that the internet was going to change the way we live and do business. Those predictions were right of course – but the way it changed our lives was heavily determined by the regulations that were put in place at the time.

In 1996 Bill Clinton, the then US president, signed the Communications Decency Act. Within that regulation was a small sub-section titled “Section 230”. Arguably, Section 230 is the key to how the whole of the internet has developed since. It states: “no provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider”. What this means in practice is that internet platforms have legal protections against being sued for the content they host.

Without Section 230, internet giants such as Facebook, Twitter, Reddit and YouTube simply wouldn’t exist. The European Union followed suit in its E-commerce Directive in the year 2000. Similar regulations with implications for other business models exist in China, for example, and have enabled the growth of Alibaba, Tencent and the like.

Read more by Hannah Ryder

• It’s time for Africa’s partners to get serious about the AfCFTA
• Why I won’t be speaking about bankability or de-risking in 2023
• How multilaterals exaggerate Africa risk

Effects across the globe

Now think about how these platforms have impacted publishing industries, e-commerce, or e-payments around the world – or even how democracy works or doesn’t work on our continent. It quickly becomes apparent how, in our globalised world, regulation from across the world can impact people on continents thousands of miles away.

And increased regulation around AI is coming.

In May this year more than 300 executives, researchers and engineers working on AI issued a statement calling for better regulation. Even Elon Musk has called for better regulation. Britain is planning an AI summit on regulation, in the hope that London will be the home for a global AI watchdog.

The issue, of course, is: in whose interests will any new regulations be crafted? What new business models will these regulations unleash and enable and what will they stifle across the world?

The fear of course is that they will act in the interests of those who have got a seat at the metaphorical table, and of those who shout the loudest. But at present these do not seem to include African businesses, still less African civil society.

These anxieties become even more ominous when Africa’s large population is cited as the key reason to “invest in Africa”. AI needs people – information, data points – to flourish.

That’s why African voices – at the very least from the AI hubs on our continent – should be represented in forming the new rules that are going to shape the future of AI.

What are the best rules for Africa?

We need to ask: what rules and regulations on AI are best for us? Which rules and regulations recognise and advance some unique features of African markets – such as our progress on financial inclusion through digital technology, or the need for more jobs on the continent? What intellectual property (IP) – including our trademarks, copyright and business models – must we fiercely protect?

Indeed, the idea of a new international watchdog overseeing this new technology could be very useful.

But given Africa’s future role in global markets, there is a strong case for this new body to be based in Cairo, Lagos, Nairobi or Cape Town, rather than in London. Indeed, if it is not, there is a real risk that the digital inequalities between Africa and the rest of the world will be exacerbated by AI, not reduced or managed.

The time is now to be thinking these issues through – and planning ahead – before AI really scales up on the continent, because by then, it will be too late to pull back.