In the age of advanced algorithms, flashing a classic peace sign for the camera is no longer just innocent fun. Asian media are sounding the alarm over the growing threat of biometric crime risks directly linked to photos uploaded to social media.
The scale of the problem was recently exposed by Chinese security expert Li Chang during a live television broadcast. He demonstrated how artificial intelligence tools can efficiently extract and clone fingerprints from a celebrity selfie.
“If the pads of the fingers are directly exposed toward the camera and photographed from within about 1.5 meters of the lens, there is a high possibility that fingerprint information can be extracted relatively clearly” – Chang warned.
The researcher added that even at a distance of 1.5 to 3 meters, today’s algorithms are still capable of recreating roughly half of a fingerprint’s key details.
These reports caused a massive stir in South Korea, where locals are famous for heavily using distinctive hand gestures in front of the lens. The traditional “V” sign, crossed fingers forming “finger hearts,” or resting a chin in a “flower pose” are absolute staples of group photos there.
Similar discussions are rippling across the Western web. On the popular tech forum Reddit, netizens are reacting with a mix of dark humor and panic. Other users, however, rightly point out that the white hat hacker community – led by German researcher Jan Krissler – was warning about these types of attacks over a decade ago. Today’s exponential leap in AI has simply automated and accelerated the entire process.
Industry experts advise extreme caution but are trying not to fuel the panic. Pei Zhiyong of the Qianxin Industry Security Research Centre reassured the media that a successful theft still requires perfect lighting, the right angle, and near-flawless lens focus. At the same time, he offers a simple, universal piece of advice for all internet users:
“It’s safer not to share too many photos showing your fingers online and to avoid saving your fingerprints on devices you don’t fully trust”.