Sensity AI gets EIC funding for forensics-focused deepfake detection models

Sensity AI , a biometric deepfake detection firm based in Amsterdam, has emerged in the top 2 percent of European startups selected to receive funding through the European Innovation Council (EIC)’s accelerator evaluation round, according to a company release. The company will receive a total of €4.9 million (about $5.6 million) – including €2.5 milion ($2.85 million) as an equity-free grant, which Sensity says “will be used in part to further develop foundational models for deepfake detection and optimize real-time deepfake analysis in resource-constrained environments.” The EIC, the EU’s flagship funding program for cutting-edge tech startups, selected 38 companies to receive funding, from an initial total of 1,760 proposals. Sensity says its product was “the only cybersecurity solution to make the grade.” In celebrating the funding, the company says it will accelerate its international expansion plans – but emphasizes its ties and commitment to Europe. “Sensity AI is a proudly European company and we are grateful for this opportunity to receive blended funding from the EIC to put our technology on the global stage,” says Giorgio Patrini, the firm’s CEO. “Furthermore, the support from the EU sends a strong message reaffirming tech sovereignty in a critical security sector with geopolitical implications.” “The deepfake threat is affecting multiple critical sectors, with 75 percent of all businesses experiencing deepfake fraud in 2025. There is an urgent need for lasting, effective solutions that can not only detect synthetic media in real-time, but can also be trusted with their forensic validity by law enforcement and intelligence professionals,” said Giorgio Patrini, CEO and co-founder of Sensity AI. The EIC investment follows the closure, earlier this year, of a $2.1 million funding round, led by Auriga Cyber Ventures with participation from Betaworks Ventures and a group of European angels. The firm is targeting law enforcement , forensics labs and governments among customers for what it calls “a validated crime-busting tool with a forensic engine producing reports for evidentiary use.” The forensic element is what differentiates Sentity in the deepfake detection market – a response to the problem of obsolescence in deployment. “All deepfake detection platforms on the market share a critical limitation, where classifiers trained on specific generators lose accuracy when encountering synthetic media from generators not represented in their training data,” the company says. Binary classification “incentivises detection models to learn generator-specific artifacts, resulting in detectors that are brittle and require costly manual retraining every time a new generation technique gains traction.” In other words, most detection models are only as good as the data they train on. Toss a new variety of synthetic media into the mix, and they’re like cars: instantly devalued. Sensity says its detection tech is “capable of learning fundamental forensic patterns,” and “decouples the learning of forensic features from the classification of specific generators,” making it more resilient to ongoing evolutions in deepfake technology.

Sensity AI, a biometric deepfake detection firm based in Amsterdam, has emerged in the top 2 percent of European startups selected to receive funding through the European Innovation Council (EIC)’s accelerator evaluation round, according to a company release. The company will receive a total of €4.9 million (about $5.6 million) – including €2.5 milion ($2.85 million) as an equity-free grant, which Sensity says “will be used in part to further develop foundational models for deepfake detection and optimize real-time deepfake analysis in resource-constrained environments.”The EIC, the EU’s flagship funding program for cutting-edge tech startups, selected 38 companies to receive funding, from an initial total of 1,760 proposals.

Sensity says its product was “the only cybersecurity solution to make the grade.”In celebrating the funding, the company says it will accelerate its international expansion plans – but emphasizes its ties and commitment to Europe. “Sensity AI is a proudly European company and we are grateful for this opportunity to receive blended funding from the EIC to put our technology on the global stage,” says Giorgio Patrini, the firm’s CEO.

“Furthermore, the support from the EU sends a strong message reaffirming tech sovereignty in a critical security sector with geopolitical implications. ”“The deepfake threat is affecting multiple critical sectors, with 75 percent of all businesses experiencing deepfake fraud in 2025.

There is an urgent need for lasting, effective solutions that can not only detect synthetic media in real-time, but can also be trusted with their forensic validity by law enforcement and intelligence professionals,” said Giorgio Patrini, CEO and co-founder of Sensity AI.The EIC investment follows the closure, earlier this year, of a $2.1 million funding round, led by Auriga Cyber Ventures with participation from Betaworks Ventures and a group of European angels. The firm is targeting law enforcement, forensics labs and governments among customers for what it calls “a validated crime-busting tool with a forensic engine producing reports for evidentiary use.

”The forensic element is what differentiates Sentity in the deepfake detection market – a response to the problem of obsolescence in deployment. “All deepfake detection platforms on the market share a critical limitation, where classifiers trained on specific generators lose accuracy when encountering synthetic media from generators not represented in their training data,” the company says. Binary classification “incentivises detection models to learn generator-specific artifacts, resulting in detectors that are brittle and require costly manual retraining every time a new generation technique gains traction.

”In other words, most detection models are only as good as the data they train on. Toss a new variety of synthetic media into the mix, and they’re like cars: instantly devalued.

Sensity says its detection tech is “capable of learning fundamental forensic patterns,” and “decouples the learning of forensic features from the classification of specific generators,” making it more resilient to ongoing evolutions in deepfake technology.