Use of AI surveillance is growing around the world

When you think of nations using artificial intelligence (AI) -enhanced surveillance technologies, China probably comes to mind: the place where facial recognition is used to ration toilet paper, to name and shame jaywalkers, and to outfit police with glasses to help them find suspects.

It’s not just China, of course. According to a report from the Carnegie Endowment for International Peace, the use of AI surveillance technologies is spreading faster, to a wider range of countries, than experts have commonly understood.

The report found that at least 75 out of 176 countries globally are actively using AI technologies for surveillance purposes, including smart city/safe city platforms, now in use by 56 countries; facial recognition systems, being used by 64 countries; and smart policing, now used by law enforcement in 52 countries.

The report’s author, Steven Feldstein, told AP that he was surprised by how many democratic governments in Europe and elsewhere – just over half – are installing AI surveillance such as facial recognition, automated border controls and algorithmic tools to predict when crimes might occur:

I thought it would be most centered in the Gulf States or countries in China’s orbit.

The report didn’t differentiate between lawful uses of AI surveillance, those that violate human rights, or those that fall into what Feldstein called the “murky middle ground.”

Smart city technologies are an example of how murky things can get. In Quayside, the smart city that’s developing on Toronto’s eastern waterfront, good intentions come in the form of sensors meant to serve the public, in that they’re meant to “disrupt everything,” from traffic congestion, healthcare, housing, zoning regulations, to greenhouse-gas emissions and more. But Quayside is also referred to as a privacy dystopia in the making.

The purpose of the research is to show how new surveillance technologies such as these are transforming the way that governments are monitoring and tracking us. It tackles these questions…

• Which countries are adopting AI surveillance technology?
• What specific types of AI surveillance are governments deploying?
• Which countries and companies are supplying this technology?

The Carnegie Endowment for International Peace presents the answers in the first-ever compilation of such data, which it’s calling the AI Global Surveillance (AIGS) index: a “country-by-country snapshot of AI tech surveillance”, mostly concerned with data pulled in from 2017 to 2019. Here’s the full index.

Some highlights:

China doesn’t just use a lot of AI surveillance. It’s also a big exporter of the technologies. The research found that Chinese companies – particularly Huawei, Hikvision, Dahua, and ZTE – supply AI surveillance technology to 63 countries. Huawei alone is responsible for providing AI surveillance technology to at least 50 countries worldwide. “No other company comes close,” the report says. The next largest non-Chinese supplier is Japan’s NEC, which supplies AI surveillance tech to 14 countries.

Chinese vendors often sweeten their product pitches with offers of soft loans to encourage governments to buy. That works particularly well in countries with underdeveloped technology infrastructures, including Kenya, Laos, Mongolia, Uganda, and Uzbekistan, which likely wouldn’t be able to get the technology otherwise. From the report:

This raises troubling questions about the extent to which the Chinese government is subsidizing the purchase of advanced repressive technology.

US companies are also active in worldwide exports. 32 countries are getting their AI surveillance technologies from the US. The most significant exporters are IBM, selling to 11 countries; Palantir, selling to 9; and Cisco, selling to 6.

Other companies based in liberal democracies are proliferating the technologies. France, Germany, Israel, and Japan aren’t “taking adequate steps to monitor and control the spread of sophisticated technologies linked to a range of violations,” the report found.