A new AI system has been developed to address the global issue of shoplifting. Vaakeye, created by Japanese startup Vaak, uses artificial intelligence to detect shoplifters based on their body language and notifies store employees through an app to prevent theft.
The system utilizes footage from security cameras to distinguish between honest shoppers and potential shoplifters based on body language cues such as fidgeting and restlessness. Vaak CEO Ryo Tanaka explained that the AI system identifies suspicious behavior using over 100 aspects of shoppers’ actions, including their gait, hand movements, clothing choices, and facial expressions.
In 2018, the software reportedly helped catch a shoplifter at a convenience store in Japan, marking a significant milestone for Tanaka. He expressed his conviction that AI can contribute to preventing crime, which is evidenced by global retailers suffering an estimated $34 billion in sales losses due to shoplifting in 2017, according to Bloomberg Quint.
The Tokyo-based company has begun selling the shoplifting-detection system and aims to have it implemented in 100,000 stores across Japan within three years. Retail analyst Thomas O’Connor from Gartner noted that many US and UK retailers already have CCTV cameras in their stores and integrating them with an analytics tool like Vaakeye can significantly improve the efficiency and effectiveness of theft prevention.
However, a study by MIT and Stanford University revealed that commercial AI systems can exhibit biased behavior, including racism and sexism, based on the data they are trained on. This has raised concerns, as seen with Amazon’s AI-based facial recognition system facing criticism for biased results. Despite this, Tanaka emphasized that his company is only beginning to realize the market potential of their software.
Image Source: Andrey_Popov / Shutterstock