Sands Lab (411080.KQ), an AI security company, showcased on-device AI forgery and manipulation detection technology at a global technology expo.
The company said on the 16th that it ran a hands-on exhibition program at Tech Week Singapore 2026, held Sept. 29-30 at Marina Bay Sands in Singapore.
The appearance was part of the 2026 project to demonstrate and spread technologies countering the adverse effects of AI, run by the Ministry of Science and ICT and the Korea Internet & Security Agency (KISA). Sands Lab is developing on-device detection and analysis technology to counter AI misuse, including authentication bypass during electronic know-your-customer (eKYC) checks and digital fraud.
Tech Week Singapore is a major Asian technology exhibition spanning information and communications technology fields such as cloud and AI infrastructure, cybersecurity, big data and AI, and data centers. Sands Lab, attending for the first time this year, introduced its technology for countering the adverse effects of AI to overseas visitors at the big data and AI exhibition.
At the core is technology that analyzes whether an image has been forged or manipulated directly on the user's device, without routing it through an external server. Sands Lab has been advancing its on-device technology based on earlier research and development results, and demonstrated Fakecheck, which applies the technology, at the exhibition.
Fakecheck is a solution that detects signs of forgery or manipulation in facial images and presents the basis for its judgment. It shows which part of the face was analyzed and why manipulation is suspected, giving users a reference for their final judgment.
On PCs, the company demonstrated a browser extension that automatically inspects images embedded in a webpage and displays suspicious items as a list. Selecting an item from the list moves the user to where that image appears for direct review.
On mobile devices, the company demonstrated a live screen detection feature alongside image forgery analysis. It analyzes faces displayed on the screen even while another app is in use and marks areas suspected of manipulation with a yellow box.
A key feature is that original images are not transmitted to an external server during analysis, reducing the risk of personal data leaks. The tool works without an internet connection and supports three languages: Korean, English and Chinese. Sands Lab is also developing technology to extend detection beyond facial images to video.
"As AI use spreads worldwide, technology that counters the adverse effects of AI, such as authentication bypass and digital fraud, and raises trust in digital content is becoming more important," said Choi Min-jun, a project manager at Sands Lab. "We will continue to contribute to the use and spread of technologies countering the adverse effects of AI through technological advancement and demonstration."