“One of our strengths is a fast FHE library, which is 30+X faster than competitors from our end-to-end optimization. We provide our product with numbers for a large dataset. It is hard to find numbers from others”
Governing bodies across the globe are supporting and investing in technology-powered initiatives. The initiative requires reliance on a technology platform.
“Digital Platform Government” is the catchphrase of the Korean government. This approach exposes many types of public data, such as taxes, education, and transportation. The scenario emphasizes the need for data privacy solutions.
“Since privacy protection is crucial for such services, CryptoLab is collaborating with the Korean Government to provide a FHE-based solution to raise the bar of privacy protection. Statistics Korea announced its plan to launch a FHE-based statistics service by 2028,” says Jung Hee Cheon, founder and CEO.
Also, the healthcare industry sees an increasing need for sensitive medical information. Cyber criminals have been targeting such sensitive information. Prevention of personal or sensitive data loss is an area of interest for CryptoLab.
During the COVID-19 pandemic, identifying contact with confirmed cases was important but raised privacy concerns. CryptoLab partnered with the Korea Disease Control and Prevention Agency to launch Cordong-I. The app uses digital contact tracing to inform users of their contact status with confirmed cases, safely protecting location information with Homomorphic Encryption. Users can use the government system without privacy concerns. “Cordong-I” is the first commercialized mobile app to use homomorphic encryption technology, making it a significant breakthrough in the field.
CryptoLab has collaborated with general hospitals and healthcare companies. The firm has partnered with Macrogen, a company specializing in genetic and genomic analyses, for a safe AI service that predicts diseases while protecting personal information.
The finance industry has a high demand for data collaboration. Single banking institutions may find it tough to detect fraud and know customer credits. FHE enables financial institutions to use heterogeneous data while minimizing the risk of personal information leakage. CryptoLab is closely working with several banks in Korea and the US. These collaborations have enabled the firm to complete many proof-of-concept projects.
Also, as AI becomes popular, privacy concerns are increasing. Many compliance rules, such as GDPR and CCPA require privacy protection by design. FHE is a good solution for using AI without risk. CryptoLab will open a new service named AutoFHE, a privacy-preserving AutoML service. The service ensures anyone can use AutoML service without concern about data leakage. CryptoLab is working on privacy-preserving generative AI.
CryptoLab is a global leader in HE. The company was initiated by Jung Hee Cheon, who invented the 4th generation fully homomorphic encryption scheme, named CKKS. CKKS is the only FHE scheme supporting real numbers needed for AI/ML.
In 2021, CryptoLab released Cordong-I, the first mobile app in the world to utilize HE, and more recent work by CryptoLab demonstrates that FHE may be used effectively for AI training. Additionally, the world’s new FHE-based SaaS service, AutoFHE, will be available shortly. CryptoLab’s product HEaaN enables FHE-based privacy-preserving data analysis for pictures, text, and tabular data.
The founder and CEO, Jung Hee Cheon is an inventor of CKKS and also is an IACR fellow. Chief Scientist Damien Stehlé is the author of two NIST PQC algorithms Kyber and Dilithium. More than 20 papers were published by the CryptoLab team in top conferences in the last five years.
Also, the firm retains a unique focus on fundamental FHE technology and applications, unlike many of its peers. “One of our strengths is a fast FHE library, which is 30+X faster than competitors from our end-to-end optimization. We provide our product with numbers for a large dataset.It is hard to find numbers from others,” continues Jung Hee Cheon.
CryptoLab has expanded overseas to France and the US and developed a cloud-based image processing solution called AutoFHE to strengthen its homomorphic encryption solution. The company is also working on installing a homomorphic encryption system in Korea’s national data-sharing system to introduce a strategy for safe collaboration and analysis of raw data.
CryptoLab completed standardization work on personal information protection and utilization technology in Korea in 2019. There are plans to finish global standards such as ISO/IEC standardization by 2025. To achieve this, CryptoLab holds an annual conference with scholars, companies, and government officials from various countries to share information and cooperate.
