Recognition systems anonymity dead

IMAGE CREDIT:

iStock

Recognition systems: Anonymity is dead

Different recognition systems use hundreds of data points to identify and track us, as well as predict our future actions in the online and physical worlds.

Author:
Author name
Quantumrun Foresight
February 6, 2023

Recognition software and algorithms have enabled many companies to profit from public information, such as photos, social media posts, and website cookies. While the technology has the potential to create relevant online experiences, some critics worry that its increasing use in public surveillance is violating human rights.

Recognition context

Surveillance systems empowered by artificial intelligence (AI) are becoming indispensable across many industries. Some examples include airports, entertainment venues, stadiums, hotels, casinos, and shopping centers. According to Technology Magazine, at least 75 countries were using AI-enabled surveillance in 2019, particularly facial recognition technology (FRT).

The benefits of recognition systems are clear. They allow companies to learn more about their customers and provide them with a personalized experience that could not be achieved any other way. Recognition systems also help law enforcement agencies keep track of criminals and prevent crime. Responsive technologies understand context and interact with us accordingly, empathizing with us when we're sad and expressing enthusiasm when we're excited.

As such, many companies are focusing on developing these systems. For example, Ireland-based startup Liopa is trialing a phone application that can interpret phrases mouthed by people. VisionLabs, based in Amsterdam, claims it can tell if someone is showing anger, disgust, fear, happiness, surprise, or sadness. This technology aims to track productivity and even make hiring decisions.

However, while this development has great potential, it also comes with significant risks. According to Nigel Jones, co-founder of the Privacy Compliance Hub and former Google executive, too much reliance on this tech can also be harmful. While these tools usually succeed more in a laboratory setting, they don't work as well when transferred to the real world.

Disruptive impact

Some argue that users have not consented to be scanned when they go out in public. Even something as simple as knocking on a friend's door (if they have a Ring doorbell) could result in your image being added to a police database. Because of increasing privacy concerns, calls to ban facial recognition are increasing from human rights organizations. Following public outcry, IBM, Microsoft, and Amazon have stopped producing FRT or are only working with police forces in the US that follow specific regulations.

However, many ways to recognize a person aren't limited to facial features. An example is voice/speech recognition, which allows machines to interpret human voices from other sources. This feature makes it easier for people to use their smart devices, as these machines now have the ability to recognize background noises from human speech. One of the emerging use cases of this recognition type is automatically creating subtitles for videos and live conferences. This feature allows people with hearing impairments to keep up with the conversation.

Recognition technologies also have other helpful applications. Pattern recognition can help in fraud detection, genome sequencing, and automatic medical diagnosis. Being able to analyze historical stats gives algorithms the ability to provide more accurate forecasts, which the financial industry is also using to predict stock movements.