At a time when social networks are full of retouched images and beauty filters, a Canadian startup, Trusting Pixels , is making a splash at VivaTech, the European technology innovation show in Paris.
Its goal? To unveil manipulated digital content and offer brands a unique tool to assess the authenticity of influencers.
Thanks to an advanced artificial intelligence system, Trusting Pixels analyzes influencer accounts to detect the amount of retouched content. "We examined several profiles, particularly in the beauty sector, where filters are ubiquitous," explains Alexander Jacquet, CEO of the startup. "Each influencer retouches a different quantity of images. Brands can now make informed choices: prioritize those offering more authentic content or, conversely, focus on those who enhance their image more."
The software uses machine learning algorithms to detect visual alterations such as skin smoothing, feature modification, or the addition of digital makeup. It generates a "retouching score," helping marketing agencies better understand the authenticity of sponsored content.
However, this technology faces the increasing complexity of image manipulation tools. For Richard Windsor, founder of Radio Free Mobile, detecting content generated or altered by AI remains a major challenge: “I’ve tested these systems and they’re sometimes fooled: they mistake my personal writing for AI-generated content, and vice versa. It’s a cat-and-mouse game where the generation is constantly improving, forcing detection tools to continually evolve.”
Faced with this frantic technological race, Trusting Pixels and other startups present at VivaTech embody the hope for increased transparency on social networks.
