Lattica has officially launched from stealth with a new cloud-based Fully Homomorphic Encryption (FHE) solution aimed at enhancing AI data security. The platform, led by Dr. Rotem Tsabary of the Weizmann Institute of Science, introduces a hardware-agnostic approach to securing AI workloads across diverse computing environments.
FHE enables encrypted data processing without decryption, allowing sensitive data to remain private throughout AI operations. Despite its theoretical advantages, adoption has been hindered by performance inefficiencies. Lattica addresses this challenge through its Homomorphic Encryption Abstraction Layer (HEAL), which standardizes FHE acceleration across various hardware platforms, including GPUs, TPUs, ASICs, and FPGAs.
The HEAL layer acts as a middleware, bridging encrypted data flows between AI models and compute infrastructure. This innovation aims to unlock secure AI capabilities for industries where data sensitivity is a key concern.
Lattica’s Core Focus on Healthcare, Finance, and Government
Lattica is targeting sectors such as healthcare, finance, and government, where strict data privacy regulations often limit AI deployment. These industries require solutions that guarantee data protection without compromising model performance.
Lattica’s cloud-based architecture supports secure AI workflows while maintaining compatibility with existing infrastructures. This is particularly relevant as the adoption of AI continues to grow in regulated industries. According to Cisco’s 2025 AI Briefing, 70% of CEOs express concern over network vulnerabilities linked to AI adoption, with 34% identifying security as a major barrier.
By offering FHE as a service, Lattica enables organizations to run encrypted queries on AI models, preventing data exposure throughout the entire inference process. This approach meets the needs of institutions handling highly confidential data, while aligning with compliance standards.
Backed by Leading Investors and Industry Experts
Lattica is supported by investors including Cyber Fund’s Konstantin Lomashuk and Polygon co-founder Sandeep Nailwal. The platform is positioned to advance FHE adoption through a combination of software optimization and hardware acceleration.
Initial community feedback highlights strong industry interest. In a recent FHE-focused survey conducted by Lattica, 71% of respondents supported a hybrid approach combining software and hardware as the optimal path for FHE adoption. The company has also made live platform demonstrations available to interested users. Lattica is headquartered in Tel Aviv and continues to focus on developing secure, scalable AI solutions for global enterprises.
