Intelligent glasses and facial recognition: the border of privacy at risk | Meta Facebook | WhatsApp Meta AI | Meta AI Europe | Turtles AI

 Two Harvard students have developed a disturbing technology that combines Meta smart glasses with facial search tools, allowing access to personal data of strangers simply by looking at them. This innovation raises serious concerns about the privacy and potential abuse of such tools. 

Key points:

•  The technology makes it possible to identify strangers and collect personal data in real time.
•  Students used modified Meta smart glasses to test their invention.
•  Facial search and language patterns increase the risk of privacy violations.
•  The initiative aims to raise awareness of data protection issues.

Two Harvard University students, AnhPhu Nguyen and Caine Ardayfio, have unveiled a technology that fuses Meta Ray Ban 2 smart glasses with a facial search engine called PimEyes, creating a system that can reveal personal data such as name, address and phone number simply by looking at a person. In their project, which aims to expose privacy-related vulnerabilities, they have created a device called I-XRAY, capable of collecting information about strangers in seconds. Through a demonstration video, they showed how this system, which combines reverse facial search with a large language model, enables automatic and rapid extraction of data that previously took hours of searching. 

Nguyen pointed out that Meta smart glasses look similar to regular glasses, making this technology particularly tricky for those who might want to collect information on unsuspecting people. The implementation of this invention raised ethical questions as students conducted tests on unaware subjects in real life, revealing how easy it can be to identify individuals with access to online information. Although some results were disputed, the demonstration clarified the risks associated with such tools. 

Nguyen and Ardayfio explicitly stated that they do not intend to release the code of their technology to avoid possible abuse. Rather, their intention is to make people aware of digital privacy and encourage them to opt out of invasive search engines such as PimEyes, a tool that, despite not directly identifying individuals, provides links to their images and identifying information. They also provided instructions on how users can remove their information from search engines, helping to protect their identity.

The project has drawn attention to a crucial issue: the legal and moral implications of using facial recognition technologies. While technology companies, such as Meta and PimEyes, downplay the risks by pointing out that similar dangers exist with traditional images, the possibility of abuse remains high, especially in the United States, where privacy laws are less restrictive than in the European Union. The possibility that this technology could be used for malicious purposes, such as stalking or scams, makes the issue particularly troubling. The students said they hope that awareness about privacy issues will balance the negative impact of their findings, thus underscoring the importance of collective thinking about how to protect personal data in an era when technology is advancing by leaps and bounds.