Charlette N'Guessan Desiree is one of the leaders of AI use in Africa, saying it can open doors for previously excluded users
By Kim Harrisberg
JOHANNESBURG, Jan 29 (Thomson Reuters Foundation) - Charlette N'Guessan Desiree loved maths and science as a student in Ivory Coast, but never imagined she would one day use her problem-solving savvy to develop facial recognition technology more adept at identifying and verifying African faces.
Her company, BACE Group, hopes its artificial intelligence (AI) software will be used across the continent - helping universities to verify students for financial services, banks to sign up new clients and security firms to fight crime.
N'Guessan Desiree, 27, said there would be less suspicion of tech created "by Africans, for Africans", especially given concerns that Western-designed systems are more prone to errors when identifying non-White faces.
"It gives us trust and credibility," said N'Guessan Desiree, a software engineer.
"It's time for us as young African engineers to start working on solutions to solve our local challenges," she told the Thomson Reuters Foundation in a video call.
BACE Group uses an API - application programme interface - software that can be used in existing apps or systems, together with facial recognition technology powered by artificial intelligence (AI).
This tech helps online businesses or financial institutions to verify their customers' identities remotely, allowing users to prove in real time that they are who they say they are without traveling long distances.
BACE recently launched a consulting arm to advise African businesses on how the technology can make it easier for them to verify new customers, such as people living in rural areas.
But despite the technology's huge potential for businesses, N'Guessan Desiree acknowledged the possible pitfalls.
"It's very important to be responsible and ethical with AI," she said.
"That's why we have defined policies between businesses and end users and we make sure the client is aware of how the data will be used by the company."
While AI facial recognition is being used globally to tackle diseases or prevent cyber theft, digital rights groups say it runs the risk of being misused to violate privacy or discriminate against users.
From targeting protesters in India to incorrectly identifying African Americans as criminals, campaigners warn that the technology can be dangerous both when it works well and when it works badly.
This week, 100 businesses, governments and nonprofits launched the Global AI Action Alliance, an initiative to make artificial intelligence more ethical and transparent, at a virtual meeting of the World Economic Forum.
The alliance will identify tools and best practices to safely and ethically use AI, which is projected to contribute more than $14 trillion to the global economy by 2035.
"This is something we want to be involved in," said N'Guessan Desiree, who added that the lack of robust digital rights laws across Africa means strong company policies are necessary to prevent data abuse.
But she said technology can be "beautiful and have significant impact" when it is used for good.
"From agriculture to finance to education, there is no limit. We can innovate and solve problems using tech," she said.
BACE is already gathering lots of attention. N'Guessan Desiree won the 2020 Royal Academy of Engineering's Africa Prize for Engineering Innovation in September last year, the first woman ever to win the award.
"I want to motivate little sisters to say 'It's possible. If Charlette has done this, why not me?'," she said.
South Africa's plan to record all babies' biometrics raises privacy fears
Africa's Boris the BabyBot opens children's eyes to digital privacy
Product design to racial justice: Global plan to promote ethical AI
(Reporting by Kim Harrisberg @KimHarrisberg; Editing by Helen Popper. Please credit the Thomson Reuters Foundation, the charitable arm of Thomson Reuters, that covers the lives of people around the world who struggle to live freely or fairly. Visit http://news.trust.org)
Our Standards: The Thomson Reuters Trust Principles.