AI Surveillance in a Post-Pandemic World
Artificial intelligence (AI) has made waves across industries including automotive, health care, entertainment, finance, manufacturing, and more, while still a nascent technology. Another area where it made a significant impact is law enforcement (with AI surveillance).
AI is now playing a pivotal role in real-time surveillance. AI surveillance helps law enforcement agencies achieve better security through prevention. In this scenario, AI-powered cameras don’t just watch the public; they also engage in intelligent analysis.
If this reminds you of “Minority Report,” you’ve hit the nail on the head. While the Hollywood blockbuster took AI and surveillance to another level, we’re not that far off.
As citizens, AI, along with deep learning (DL) and machine learning (ML), will have a profound impact on us all. This is evidenced by research that suggests that the global video surveillance market is worth about $45.5 billion now, growing at a CAGR of 10.4% to $74.6 billion by 2025.
AI Surveillance Explained
AI surveillance describes the process of ML and DL-based algorithms analyzing images, videos and data recorded from CCTV cameras. The most common form of AI-based monitoring is rule-based. This is a process in which software engineers feed the system with pre-designed rules that must be true in an ideal situation.
However, as it’s a self-learning system, behavioral analytics is used to auto-detect and classify data related to the target environment. In this scenario, when trained machine-based vision detects the slightest deviation from normal activity, the authorities are notified immediately.
Depending on parameters including accuracy, speed and size, popular approaches include the following:
- Faster Region-CNN (RCNN)
- Single Shot Detector (SSD)
- You Only Look Once (YOLO)
Although still nascent, facial recognition is now advanced enough to leverage both still images and video footage to confirm a person’s identity. They do this through database matching, expression analysis through crowd scanning and comparisons with other similar trends.
AI Surveillance: Not Just for China
China has leveraged AI-powered surveillance technologies for years. In fact, the authoritative regime has used it as a tool to score people on social behavior and even punish them.
Such misuse of technology to take away freedoms and oppress people gave us all cause for concern. The obvious threat is to human rights, but there are also broader implications.
Real-time facial recognition startup Clearview AI, for example, got caught scraping images from social media profiles. Cease-and-desist letters quickly were issued by leading social media platforms, including Facebook, Twitter and LinkedIn, as well as Google.
As Clearview’s technology shares the same shortcomings of other facial recognition systems, it’s not as good at recognizing brown and black faces as it is for white. Hence, it’s safe to say that it’s not ready for policing.
In a highly racially charged environment (with minorities experiencing police brutality) in North America, the consequences threaten to be horrific. As Clearview’s security protocols are untested and unregulated, it could also compromise data privacy and security regulations.
To mitigate risk, the European Commission plans to strictly regulate AI, including a ban on “black box” AI systems (which humans can’t interpret). The primary objective here is to build the public’s trust and not induce fear.
At the end of the day, when used properly, AI promises to be a massive benefit for society.
Tracking People in a Post-Pandemic World
With the onset of the COVID-19 pandemic, societal fears surrounding AI-powered surveillance have taken a back seat. AI’s potential for helping us effectively combat COVID-19 is difficult to ignore.
European governments, for example, backed the use of contact tracing applications. Others championed the use of smart surveillance technologies to help contain the virus.
After the novel coronavirus spread outside Wuhan Province, Chinese AI giants Baidu and Megvii deployed AI-powered temperature-scanning technologies. These include scanners that detect body temperature and send alerts to the relevant authorities to help contain the virus.
In France, AI monitors the Paris metro system to ensure that passengers are wearing face masks. Created by French startup DatakaLab, the technology’s goal isn’t to identify and punish people who fail to follow healthcare guidance; instead, it aims to generate anonymous statistical data that helps authorities anticipate future outbreaks.
AI surveillance during the pandemic can help countries stay a step ahead of the invisible virus. So even if you have doubts about the technology or fear its potential for weaponization, the benefits are difficult to ignore.
Unless there’s a vaccine, we can expect AI-powered surveillance in our offices, malls and public places to keep us safe. In the meantime, robust legislation is necessary across borders to protect privacy, security and human rights.
Regardless of how you feel about AI surveillance and tracking, this technology is here to stay. Whether it’s in health care or policing, when used correctly, the benefits far outweigh its potential shortcomings.
