Lattica emerges from stealth with $3.25M to make encrypted AI a reality using FHE

AI adoption has grown quickly since ChatGPT’s breakout two years ago, but trust hasn’t kept pace. While businesses are eager to explore AI, concerns around data privacy and security are slowing progress, especially in industries that handle sensitive information. According to Cisco’s 2025 AI Briefing: CEO Edition, 70% of CEOs remain uneasy about the security of their networks, and 34% see it as one of the biggest hurdles to AI deployment.

Lattica, a Tel Aviv-based startup, is tackling this problem head-on with Fully Homomorphic Encryption (FHE)—a breakthrough encryption technique that allows computations to be performed directly on encrypted data, so it never has to be exposed.

Startup Lattica Raises $3.25M to Solve AI Security with Fully Homomorphic Encryption

Fresh out of stealth, Lattica has raised $3.25 million in pre-seed funding to bring encrypted AI to the cloud and prove that FHE is no longer just a theoretical concept—it’s ready for real-world use.

The round was led by Konstantin Lomashuk’s Cyber Fund, with backing from several notable investors, including Polygon Network co-founder Sandeep Nailwal and Sentient: The Open AGI Foundation.

FHE isn’t new—it’s been called the “holy grail” of cryptography for years. The tech makes it possible to process encrypted data without ever decrypting it. The catch? It’s been painfully slow and resource-hungry, far from practical for AI. Lattica believes it has cracked that problem by pairing FHE with modern AI acceleration techniques.

At the center of its solution is a platform that lets users run AI queries securely through a hardware-agnostic cloud. That means the data stays encrypted during the entire process—training and inference included.

Dr. Rotem Tsabary, the founder and CEO, brings academic firepower to the effort. She holds a PhD in lattice-based cryptography from the Weizmann Institute and sees parallels between machine learning operations and FHE. That overlap gave her team the edge to make FHE commercially viable for AI.

One of Lattica’s core innovations is something it calls HEAL—short for Homomorphic Encryption Abstraction Layer. It acts as a connective layer between FHE applications and AI algorithms, and works across GPUs, TPUs, CPUs, and even custom chips like ASICs and FPGAs. Think of it as infrastructure that finally makes encrypted AI fast enough to use in real-world settings.

“By combining the advancements of hardware acceleration with software-based optimization, we realized that not only could we improve FHE efficiency to the point of commercial viability, but use it to solve critical data dilemmas holding back AI’s adoption in sensitive industries,“ said Dr. Rotem Tsabary.

The company is focusing first on healthcare and financial services—two sectors where privacy laws and regulatory headaches have made it hard to put AI to use. From secure data analysis in clinical trials to encrypted financial models, Lattica is betting that FHE can move these industries past their trust issues with AI.

As part of its launch, the team released demos on its site, along with results from a survey it ran across the FHE community. The big takeaway: 71% of respondents believe that the key to FHE adoption lies in combining hardware and software breakthroughs. That lines up neatly with what Lattica is building.

“Lattica is pushing the boundaries of Fully Homomorphic Encryption, solving one of the most critical challenges in AI security,” said Konstantin Lomashuk, Managing Partner at Cyber Fund. “Cyber Fund is proud to have led Lattica’s pre-seed round. This is the kind of deep-tech innovation that defines the future, and we’re excited to see Lattica leading the way.”

Sandeep Nailwal added, “Lattica’s product-first approach fundamentally transforms sensitive data processing in the AI ecosystem. Lattica has made FHE a reality that is both practical and scalable, as Tsabary and her research team is proving that advances in the machine learning stack can significantly boost the performance of FHE and have an immediate impact on the market.”

With its cloud-based service and encryption-first approach, Lattica wants to make querying an AI model as safe as sending a secure message—no decryption required, no trust needed.