Eleven African Governments Built a $2 Billion Chinese Surveillance Grid

Imagine a company that sells you a security camera for your home, lends you the money to buy it, and then gets to keep a copy of everything the camera records. That is roughly the deal that 11 African governments signed up for, and the total bill has now crossed $2 billion. And, of course, it only gets worse.

A report published on March 12 by the Institute of Development Studies and the African Digital Rights Network documents the spread of Chinese AI surveillance across Algeria, Egypt, Kenya, Mauritius, Mozambique, Nigeria, Rwanda, Senegal, Uganda, Zambia, and Zimbabwe. It covers facial recognition, vehicle tracking, biometric collection, and centralized command infrastructure, with almost all of it being financed through Chinese state bank loans.

Facial Recognition Bankrolled by Beijing

The technology packages in question follow a very familiar template. Chinese companies, primarily Huawei and ZTE, bundle thousands of CCTV cameras with facial recognition and automatic number plate recognition capabilities, connect them to a central surveillance command room, and finance the whole thing through loans from Chinese state banks.

Nigeria has spent the most, pouring $470 million into 10,000 smart cameras. Mauritius follows at $456 million, Kenya at $219 million, and Egypt, Algeria, and Uganda each operate around 5,000 to 6,000 cameras. The average investment across all 11 countries sits at $240 million, much of it borrowed from the same Chinese institutions supplying the technology.

The pitch, naturally, is simple: modernize your cities, reduce crime, and look like a 21st-century government. That is, if by 21st-century government you mean only all the worst sides of it, and by modern cities you mean exactly the same scene, only with a bunch of cameras on every corner.

And as for the crime reduction argument, guess what? It’s pure fiction. Researchers found no credible evidence that any of these systems have actually lowered crime rates in the countries that deployed them.

The Real Targets Are Not Criminals

The researchers are quite direct about what these systems are actually doing. Governments across the 11 countries are using surveillance infrastructure to monitor human rights activists, track political opponents, arrest protesters, and create conditions where journalists start self-censoring before anyone even knocks on their door.

Report co-author Wairagala Wakabi, executive director of the Kampala-based policy body CIPESA, said the scale of deployment cannot be justified on security grounds, describing it as surveillance that is not "legal, necessary, or proportionate" to any legitimate aim. I'd say that's a remarkably restrained way of describing a $2 billion operation with no meaningful legal framework governing it in any of the 11 countries involved.

Georgetown professor Bulelani Jili added a warning that even introducing legal frameworks at this point could backfire, since surveillance laws have a track record of ending up as tools that criminalize dissent rather than regulate the technology. And in Zimbabwe, the situation has an extra dimension, as the government's deal with Chinese firm CloudWalk ships biometric data back to China for facial recognition algorithm training, which researchers have described as a form of data colonialism.

This is exactly the pattern of surveillance normalization that has played out in country after country, and Gabon's government showed how quickly even more basic tools, like social media kill switches, get turned on citizens the moment political pressure rises. Now consider what something of this scale could do.

Infrastructure Built to Outlast Any One Government

The report also traces a direct line from colonial-era intelligence networks to today's AI-enabled monitoring systems, and that historical framing matters more than it might seem. Surveillance infrastructure does not get dismantled when administrations change. It gets inherited, expanded, and eventually normalized.

Once a network of 10,000 facial recognition cameras exists, it gets used, and the people most likely to end up in its crosshairs are not the ones governments claim to be protecting citizens from. The $2 billion spent across these 11 countries was not an investment in public safety but a way to keep governments safe from their own people. The evidence for that is in who gets watched, with the list mainly containing activists, journalists, and protesters, not the actual sources of crime these systems were supposedly built to address.

If any of these governments want to claim otherwise, the burden of proof is entirely on them to show at least a single credible study demonstrating crime reduction, an independent oversight body reviewing how the data is used, and a legal framework with real teeth protecting the people being filmed. Until then, $2 billion worth of cameras is just $2 billion worth of excuses.