North American researcher Nathaniel Allen told Lusa that the Artificial intelligence (IA) “is developing so quickly that African governments do not – nor does anyone else – have the legislative frameworks to deal with some of its potential threats".
“When we have AI-driven algorithms that can bypass verification systems and create convincing deep fakes of world leaders, it's difficult to know how to respond, and responses typically take time,” added Nate Allen, researcher and professor of Security Studies at the Africa Center for Strategic Studies (ACSS), in Washington.
“Many African governments will face major challenges in security and in enforcing any regulatory frameworks that provide a common ethical basis by which governments, businesses and citizens can be held accountable,” he said.
What happens with AI, he further illustrated, is currently happening with the attempt to classify cybercrime worldwide.
“You cannot apply a law if you do not know what it applies to, and this is the big problem with cybercrime, in which the world cannot agree on what constitutes it and there are major disputes between authoritarian and democratic regimes over a global treaty that prevents it, cybercrime as such is understood globally”, explained the ACSS researcher.
On the other hand, he highlighted, the simple fact that there is a law “can make a huge difference”, however, “just as important or more important than having guidelines, is that countries consider the impact of AI on the laws that are already in force” .
BUT: OpenAI launches artificial intelligence that turns text into video
For example, the expert asked, “what do AI-driven 'deep fakes' mean for fraud and impersonation laws and defamation laws? And how these laws may need to be updated to take into account the potential for using AI for fraud purposes.”
“This is as important, if not more important, than having some kind of comprehensive regulation on AI in place,” he concluded.
In developing policies for the use of AI in Africa, Allen said, “the focus is a little more on how this technology is incorporated for various development-related purposes and has not focused enough on its potential harms.”
Normally, it is the ministries of communication and others of the kind that deal with the problem, but, on the other hand, “there is almost no effort to reflect on the military and security implications for African countries, with regard to AI, despite AI has already been incorporated into some military technologies for some time,” he said.
“Whenever a drone uses — and drones are proliferating — AI in image recognition systems, from reading license plates to biometric data” there are ethical questions, but not only, that arise, said Allen, drawing attention to the fact that drones incorporating AI are “increasingly being used to define military targets”.
“According to reports, the first autonomous weapon, incorporating AI, was used on African soil in 2020, during the conflict in Libya. In addition to adopting technology, I am not aware of any African government that is really thinking about how to ethically use and employ these types of advances, which is a big problem”, he stated.
Nate Allen considered that the main way in which AI is having a “strategic impact” in Africa is in the way it “shapes the content that people interact with online”.
“This is an important concern, because, although there are some countries in Africa where only 20% of the population has access to the Internet, conversations on the Internet influence what happens 'offline' even if people are not aware of it,” he said. .
The question that arises, highlights Nate Allen, is: “What if this content is being influenced by a foreign actor who is spreading disinformation? Or by an algorithm designed by a social media company that aims to optimize engagement rather than being aware of the fact that it is a vehicle for fact-based political discourse.”
“The consequences are problematic, destabilizing and, in particular, countries where there is not much capacity and experience in the State to deal with this, will have problems”, defended the ACSS researcher.
Technology progresses ahead of regulators and “will always be ahead of most of their ability to understand, let alone do anything,” Allen said.
But, to answer these types of questions, added the researcher, “it is not even necessary to create solid regulation, it is enough to be aware that there is a problem in the first place”, in Africa, as in any part of the world.
