The West might have developed the initial technologies for most forms of artificial intelligence (AI), but it is in China where the AI technologies have seen massive deployment on a scale that boggles the mind. Remember the Alibaba Citi Brain project in China which I wrote about in this column on 5 November 2018? The project reportedly started in October 2016 in the city of Hangzhou in Zhejiang province where Alibaba’s headquarters is located. The project has the objective of creating a cloud-based system where information about a city, including all its inhabitants, is stored and used to control the city.
A few capabilities of Alibaba’s City Brain include: a) Ability to see every car in the city of Hangzhou, and constantly monitor video footage of traffic, looking out for signs of collisions or accidents in order to alert the police, b) ability to cover all 420 square kilometers of urban areas in greater Hangzhou city, c) optimization of the operation of 1,300 traffic lights at intersections and connecting 4,500 traffic monitoring video cameras. It is remarkable to know that City Brain isn’t just connected to Chinese authorities – to notify them when there’s an emergency or a crisis that needs handling, – it’s also connected to all the phones of the 9.5 million inhabitants of Hangzhou – to inform them of traffic and weather conditions in real-time. I think these are good uses of AI. However, the Chinese government reportedly also uses AI for law enforcement on a massive scale.
- Embrace technology for emergency situation, FCT Commissioner of Police charges residents
- Innovation, technology to save primary health care
That is China. The situation in the US and Western Europe regarding the use of AI, and, in particular, facial recognition, is much more tenuous. The combination of three factors – heavy investment by the Chinese government, huge and willing population, and the less stringent personal data protection regulations – helps the attainment of Chinese government’s AI objectives. Contrast this with the fact that starting from June of 2019, the city of San Francisco in the US has banned the use of the facial recognition technology (FRT), because, according to the city supervisor, Aaron Peskin, “facial recognition is “uniquely dangerous and oppressive.” The ban came after many studies had raised real concerns about civil liberties and the demonstrated instances of troubling racial and gender bias as well as the error rates. The apparent abuse of the technology by some governments in the world, for example, for racial profiling, is a matter of international concern that has also drawn international condemnation.
The Fate of the FRT in the US and in the West will probably be decided in 2021, and, because America and Western Europe take human rights violations very seriously, the outcome may not favor the proponents of FRT, which by and large are the tech companies that are developing the AI tools. In the West, individual citizens and civil right groups can challenge the actions of business entities and the government, to bring about tight regulation or even abolition of the use of FRT. In fact, this is already happening in the US. According Alfred Ng in cnet.com on 11 December 2020: “In 2020 alone, facial recognition bans popped up in cities including Jackson, Mississippi; Portland, Oregon, and Portland, Maine; and Boston, Cambridge and Springfield, Massachusetts. In December, Massachusetts lawmakers voted to make their state the first to completely ban police use of facial recognition.”
A few US senators have introduced bills to ban the use of facial recognition. They contend that “the growth of facial recognition surveillance poses a serious threat to our privacy and civil liberties, and it also disproportionately endangers Black and Brown Americans,” and that “we need to be rooting out forces of racial injustice and invasive surveillance in our society.” There are known cases of police making false arrests because of facial recognition’s racial bias.
Note that there is bipartisan support in the US senate for facial recognition regulations. The incoming Biden administration promises police reform; the facial recognition issue might be pinned onto to that. One interesting angle to the impending facial recognition regulations development model is that the regulations (for the government to adopt) are apparently being developed by the same companies that are the target of the FRT regulations. The Biden administration will probably stop this, if only for sanity sake.
As stated above, the tech companies developing the FRT are of course not in favor of an outright ban, even though some of them – Microsoft and IBM in particular, have expressed some support for regulations. Other entities may be less willing. In fact, this past week, Google reportedly fired one of its prominent AI researchers, Dr. Timnit Gebru, a Stanford University graduate who happens to be black, after she sent an email concerning biases built into Googles artificial intelligence technology and the company’s lack of progress in hiring women and minorities.
There is also Clearview AI, a very controversial facial recognition startup in New York City that is harvesting billions of photos of people all over the world from YouTube, Facebook, Twitter, etc., without the permission of the social networks, or, of course, the people whose data are being harvested. The Year 2021 will decide whether Clearview AI or entities with similar projects can continue to do this.
