On October 8, 2025, two major developments in the world of privacy technology converged to highlight how artificial intelligence (AI) and advanced cryptography are reshaping the way sensitive data is managed, especially in sectors like healthcare, finance, and public administration. In Europe, Bilbao-based startup Nymiz secured €2 million in Series A funding to bolster its AI-driven data anonymisation platform, while in South Korea, a government-backed research team at Asan Medical Center (AMC) announced a breakthrough in privacy-preserving AI for medical imaging using homomorphic encryption. Both stories underscore a growing trend: privacy is no longer just a regulatory checkbox but a foundational element in the digital infrastructure of tomorrow.
Nymiz, founded in 2020 by Óscar Villanueva Cañizares, has been on a mission to bridge the widening gap between data utility and privacy protection. With this latest investment round led by Amsterdam-based TIN Capital—and joined by Swanlaab Venture Factory, Auriga Cyber Ventures, SWG Cyber & Defence Fund, and CDTI—the company is poised to accelerate its expansion across Europe, Latin America, and the U.S. by 2026. According to TIN Capital, which recently launched its European Cyber Tech Fund V, Nymiz is a promising leader in the privacy-tech space. Bart Houlleberghs, Partner at TIN Capital, remarked, “Nymiz’s deep tech and clear product-market fit align perfectly with our mission to empower Europe’s cybersecurity and privacy innovators.”
What sets Nymiz apart from competitors like Privitar, BigID, and Privacera is its laser focus on unstructured data and its ability to automate anonymisation and pseudonymisation at scale using AI. The platform currently serves over 70 enterprise clients across finance, healthcare, legal, and public administration sectors. Its technology can process documents and databases in 102 languages, replacing personal information with synthetic data or tokens. This not only helps organizations comply with stringent privacy laws such as the General Data Protection Regulation (GDPR) in Europe and the Health Insurance Portability and Accountability Act (HIPAA) in the United States, but also allows them to extract valuable insights from data without risking the exposure of sensitive information.
“Our platform not only meets increasing regulatory demands but also anticipates the risks introduced by the growing use of generative AI,” said Óscar Villanueva, CEO and co-founder of Nymiz. He added, “This round is about building the operational foundation of the Nymiz Platform and preparing for the international visibility we’ll pursue in 2026. With the support of TIN Capital, Swanlaab, Auriga, and our other investors, we’re uniquely positioned to take our vision of Privacy as Infrastructure global.”
The idea of embedding privacy directly into digital systems—what Nymiz calls “Privacy as Infrastructure”—is gaining traction as organizations grapple with the dual challenge of leveraging AI while safeguarding individual rights. As AI technologies become more prevalent, the risks of data misuse, breaches, and regulatory penalties only grow more acute. Nymiz’s approach goes beyond mere compliance, aiming to make privacy a strategic advantage and a cornerstone of the modern data economy. As TIN Capital’s Bart Houlleberghs put it, “We’re excited to support CEO Óscar Villanueva and the team at Nymiz. This investment aligns with our mission to back Europe’s most promising cybersecurity and privacy-tech scaleups. Nymiz stands out with a strong team, clear product-market fit, and global ambition.”
Meanwhile, across the globe in South Korea, researchers at AMC have demonstrated that privacy-preserving AI is not just a theoretical ideal but a practical reality. In a study supported by the South Korean Ministry of Science and ICT, the National Research Foundation of Korea, and the National IT Industry Promotion Agency, the AMC team developed a deep learning model for diagnosing kidney disease from CT images, all while keeping patient data fully encrypted. The secret sauce? Homomorphic encryption—a cryptographic method that allows computations to be performed on encrypted data without ever decrypting it.
The research team began by building a baseline model using 12,446 unencrypted renal CT images covering normal, cystic, and tumor cases. They then adapted the model to function in an encrypted environment by replacing certain mathematical operations with ones compatible with encryption. The encrypted images were processed using the Cheon-Kim-Kim-Song (CKKS) scheme, developed by local startup CryptoLab. Unlike earlier homomorphic encryption techniques, CKKS can handle approximate operations on real and complex numbers, which are essential for deep learning in medical imaging.
Despite the technical hurdles—encryption ballooned image size by 500 times and increased computational demands—the team found that a high-performance GPU could still analyze the images within one to two minutes. Most importantly, the model achieved an impressive 97%-99% accuracy in disease classification, matching the performance of unencrypted models. According to AMC, “even with encrypted patient data, the model delivers analysis results at the same level as existing unencrypted models.”
Dr Lee Sang-wook, professor of Anesthesiology and Pain Medicine at AMC, believes this breakthrough could soon become the norm. “The new encryption-based model is set to become the standard for personal information-preserving medical image analysis once it is optimised and GPUs are further developed,” he said. Dr Seo Jun-kyo, professor from the AMC Department of Urology, echoed this optimism, noting the technology’s potential to minimize legal and ethical concerns around AI in healthcare.
This isn’t AMC’s first foray into privacy-preserving AI. Last year, the hospital demonstrated how homomorphic encryption could protect patient data used to train AI models, encrypting electronic medical records from more than 300,000 patients. The practical applications of such technology extend far beyond medicine—from secure elections to confidential financial analytics—showing just how versatile and vital privacy-preserving AI is becoming in a data-driven world.
Both the Nymiz funding milestone and AMC’s research highlight a pivotal shift: privacy is increasingly seen as a business enabler and a societal imperative, not just a box to tick for compliance. As organizations worldwide race to unlock the value of data, the ability to do so responsibly—without sacrificing privacy—may well determine who leads in the next wave of digital innovation.
It’s clear that solutions like Nymiz’s AI-powered platform and AMC’s encrypted medical imaging model are setting new benchmarks for what’s possible when privacy is treated as infrastructure. As these technologies mature and spread, the hope is that individuals, companies, and entire industries can benefit from AI’s promise without compromising on the fundamental right to privacy.
