identifAI has been recognised as a pioneer of real-time content validation by Gartner.
Read More >
Identifying the origin of content—often indistinguishable—as potentially generated by artificial intelligence systems. This has increasingly become a matter of public concern. And it is precisely the mission of IdentifAI, an innovative Milan-based start-up with ambitions for European expansion, particularly targeting the media sector as well as corporate leadership. Its goal: to prevent stories without any factual basis from being presented as real, by detecting artificially generated images and videos.
IdentifAI aims to provide every individual—citizen, consumer, political or business leader—with the ability to clearly distinguish whether what they are seeing is the product of artificial intelligence or human creativity. In this way, it seeks to ensure that emerging technologies serve the common good rather than becoming tools of destabilization, thanks to equally effective technologies designed to mitigate such risks.
Marco Ramilli, founder and president of IdentifAI, explains: “Today, thanks to generative artificial intelligence, it is possible to create content indistinguishable to the human eye. We can no longer tell if an image, a video, or even a sound or voice was created or generated artificially. These elements are crucial in influencing us—for example, convincing us to take a certain action. It is therefore just as essential to inform the end user, so they can distinguish reality. Warning them of AI-created content—that is where we step in.”
The scope of application is vast—essentially anywhere an image requires verification: “From identity documents to insurance fraud, to reputational risks such as fabricated content aimed at discrediting a corporate executive. Today, the information we rely on to make decisions may be of poor quality, fake, non-existent, or subtly altered. This is why everyone should have tools to understand. Big Tech companies have allowed anyone to present and modify their own version of reality. It is up to us—managers, students, journalists—to exercise great responsibility. The media must remain vigilant, using technology to reach the truth.”
Can results be guaranteed? “AI is, by definition, a probabilistic science. There are no certainties. But what we do is the reverse of this probabilistic activity—we call our systems ‘de-generative models,’ in contrast to the generative models we already know. Still probabilistic, but with an accuracy rate that, in June 2024, stood at 88% internationally. Today, we are very close to 94%.”
