Ant Group today unveiled SecretFlow Cloud, its cloud-based computational cryptography platform and a suite of computational cryptography solutions tailored for large language models (LLM).
Leveraging integrated software-hardware encryption technologies, these solutions enable the secure flow of encrypted data throughout the LLM hosting and inference process, protecting LLM asset integrity and data security while significantly enhancing data privacy protection . Providers can now encrypt and deploy their LLMs in cloud environments effortlessly with a single click through the encrypted hosting service, protecting their valuable assets from compromise and theft. Meanwhile, the encrypted inference service ensures effective data security and protection of trade secrets during interactions with LLM.
These solutions allow GPUs to perform computational tasks within trusted execution environments, primarily reducing the cost and performance difference between encrypted LLM and plaintext operations, providing a more cost-effective approach. The platform also incorporates additional technologies such as memory and disk encryption, ensuring end-to-end encryption and secure management of cross-domain model hosting. Additionally, a user-friendly remote attestation system enables smooth verification through web interfaces, minimizing friction for users.
SecretFlow Cloud currently supports both public and private cloud deployments and is compatible with LLMs commonly used in the global market. For example, in a public cloud installation, users can seamlessly create a new professional LLM on SecretFlow Cloud or migrate an existing LLM to the platform, with secure LLM inference services ready in just 10 minutes with a single click. In addition, users can purchase cryptographic computing resources on demand to meet their needs.
“Data is the most critical element in the application of LLMs, and cryptographic computing technology greatly enhances the utility of data in cross-sector scenarios,” noted Dr. Lenx Wei, Vice President and Director of Security Technology at Ant Group. “As LLMs evolve into specialized career fields and become essential productivity tools, cryptocomputing will play a vital role, particularly in unlocking the full value of high-quality, domain-specific datasets.”
According to Dr. Wei, while companies often deploy LLMs in private environments to address data security challenges, such methods can incur higher operating costs and compromise both efficiency and service quality, ultimately limiting the potential of LLMs. The computer encryption solutions offered by SecretFlow Cloud can effectively solve these issues.
Looking to the future, SecretFlow Cloud aims to further develop its computer encryption solutions to ensure data security throughout the lifecycle of professional LLMs, including their creation, development and service delivery.
Ant Group has been exploring privacy-preserving computing technologies since 2016 and released SecretFlow in 2022 as an open-source privacy-preserving computing framework that features a number of advanced technologies including multiparty computation, federated learning, trusted execution environments, homomorphic cryptography, and differential privacy.
Positioned at the forefront of cryptographic computing, SecretFlow Cloud represents a new generation of privacy-preserving computing technology poised to overcome security, cost, and accessibility pain points in data traffic. It offers a solution to instantly harness the value of data, similar to turning on a faucet. To date, its technologies have been adopted in industries such as insurance, agricultural finance, healthcare, public services and marketing.