How can IA remain private? The financing of $ 3.25 million from Lattica pushes completely homomorphic encryption under the spotlight
Can AI process sensitive data without ever exposing it? This is the question that the startup based in Tel Aviv Lattica tries to answer because it leaves the stealth mode with $ 3.25 million in pre-series funding. The company's mission: to resolve one of the most persistent confidentiality challenges of artificial intelligence using entirely homomorphic encryption (FHE).
Break the financing of Lattica and the interest of investors
The pre-series funding of Lattica, led by the Konstantin Lumashuk cyber-laundry, includes the participation of notable investors like Sandeep Nailwal, co-founder of Polygon Network and are: The Open Agen Foundation. The injection positions of 3.25 million dollars Lattica pre-series to evolve its cloud-based platform, which promises a secure AI calculation by allowing requests on encrypted data-without decrypting it.
Investors' interests report an increasing demand for technologies improving confidentiality in AI, in particular in industries, where compliance with data protection regulations is not negotiable. According to the Cisco AI briefing in 2025, security remains a major concern, 34% of CEOs citing it as an obstacle to a broader adoption of the AI.
What makes Lattica's approach different?
Fhe has long been hailed as the “Saint Grail” of cryptography, offering a means of calculating encrypted data. However, due to the ineffectiveness of performance, it has largely remained a theoretical solution. Lattica takes up this challenge through its homomorphic encryption abstraction layer (HEAL), which normalizes and accelerates operations in various material environments, including GPU, TPU and ASIC.
Dr. Rotem Tsabary, founder and CEO, explained, said,
“We allow practice by developing a tailor -made solution for neural networks.”
With an cryptography training based on a network of the Weizmann Institute, Tsabary's vision operates both hardware optimization and software to fill the gap between secure calculation and evolutionary AI deployment.
Industry focus: health care and finance in the reticle
The Lattica platform is particularly relevant for sectors such as health care and finance, where sensitive data management is both a regulatory and operational concern. Applications range from encrypted financial transactions to a secure analysis of medical data for research purposes.
Sandeep Nailwal commented,
“The approach of Lattica, focused on the product, fundamentally transforms the processing of sensitive data in the ECA ecosystem. The progress of the automatic learning battery considerably increases performance. ”
The startup of the startup within the community has revealed that 71% believe that adoption will depend on the combination of hardware and software – evaluating the hybrid approach of Lattica.
Market implications and final thoughts
The emergence of Lattica reflects a broader trend: the growing pressure on AI providers to guarantee data confidentiality at all costs. As regulatory environments are tightened worldwide and AI becomes still anchored in critical infrastructure, solutions like FHE can go from niche to general need.
Lattica's success will not only depend on its technology, but its ability to keep the performance promises where others have stalled. If Heal really provides the necessary acceleration, Lattica could be at the forefront of an AI revolution in terms of intimacy.
Fhe has always seemed too good to be practical. The hybrid model of Lattica is ambitious but opportune. Funding, although modest compared to AI start -up standards, could offer enough track to prove the viability. Pressure is now on Lattica to show measurable performance gains that can persuade industries for a long time the risk of AI confidentiality.
Do not forget to love and share the story!
Disclosure of acquired interests: This author is an edition of an independent contributor via our
