CryptoLab and UClone Partner to Bring First Fully Homomorphic Encryption-Powered AI Agents to Consumers

SAN FRANCISCO, April 28, 2025 /PRNewswire/ -- CryptoLab, a leader in fully homomorphic encryption cryptography solutions, and UClone, a pioneer in personalized AI, today announced a collaboration to integrate CryptoLab's Encrypted Vector Search (ES2) for Retrieval-Augmented Generation (RAG) into UClone's AI agent platform. This partnership marks the first time fully homomorphic encryption is being used to secure AI agents, ensuring unparalleled data privacy and security for users.

A groundbreaking demo that displays UClone's AI agent powered by CryptoLab's Encrypted Search will be showcased at the RSA Conference in San Francisco.

Key Highlights:

• Enhanced Privacy: CryptoLab's Encrypted Search. UClone agent to perform RAG operations on encrypted data, ensuring user data remains encrypted and private even during processing.
• First-of-its-Kind Solution: This collaboration brings the first AI agent solution using homomorphic encryption to the consumer market, setting a new standard for data privacy in AI applications.
• Private and Personalized AI: By leveraging homomorphic encryption, UClone agents can deliver personalized experiences while maintaining the highest levels of data security and compliance.
• RSA Conference Demonstration: Attendees of the RSA Conference in San Francisco will have the opportunity to witness the live demonstration of this innovative technology.

"This is a pivotal moment for data security and private AI," said Michael Pak, CPO of CryptoLab. "It is fitting that the UClone–CryptoLab partnership launches at RSA, as two of the pioneers behind RSA cryptography—Rivest and Adleman—along with Dertouzos, first proposed the concept of homomorphic encryption in 1978." Pak added, "Through the half-century of cybersecurity, many technologies have come and gone, but cryptography has remained the security anchor for information technology innovations." He further stated, "The partnership between UClone and CryptoLab marks the beginning of a new era; one where the convenience and power of AI are backed by data privacy, anchored in mathematics."

"Privacy and security are front and center at UClone. Our users trust us with sensitive information, and we chose CryptoLab's Encrypted Search for its unmatched data privacy, seamless integration, and resilience against both current and future cyber threats," said Kenny Lim, Advisor at UClone. Lim added, "As we expand our groundbreaking multi-agent AI social platform to new customers and markets, we look forward to a lasting partnership with CryptoLab and broader adoption of homomorphic encryption technology."

About CryptoLab:

CryptoLab, Inc. (www.cryptolab.co.kr) is a leader in fully homomorphic encryption (FHE) technology. We hold the original patent for the widely adopted 4th generation FHE scheme CKKS. Our mission is to deliver practical yet mathematically proven data privacy solutions that enable private AI.

At this year's RSA Conference, CryptoLab will showcase Encrypted Vector Search, a high-performance data privacy system that enables the processing of sensitive vector embeddings, such as those used in facial recognition and personalized RAG-LLMs without decryption. CryptoLab is headquartered in Seoul, Korea, with offices in France, and San Jose, California. For more information, please visit us at RSA booth S-0634 or email us at [email protected].

About UClone:

UClone is a Silicon Valley startup at the forefront of the artificial intelligence revolution, offering the groundbreaking UClone service. UClone is a multi-agent AI social platform where each AI Clone is an agent residing in a virtual environment. These agents interact and collaborate, accumulating knowledge through their interactions. UClone focuses on enabling private data handling and secure communication between agents, incorporating private RAG (Retrieval-Augmented Generation). UClone is operated by UClone Inc., a Silicon Valley startup founded in 2024. (http://www.uclone.net).

SOURCE CryptoLab