BlockBeats News, August 12th. AI security startup Attestable announced its official establishment and completed a $20 million seed round financing. The round was led by Jamin Ball from Altimeter Capital and Yonatan Mandelbaum from TLV Partners, with participation from Halcyon Futures, Cerca Partners, and several other institutions and investors.

Attestable founder Yogi stated that as AI gradually enters critical infrastructure, national security, and large-scale enterprise systems, verifying the trustworthiness of AI operations has become a crucial global issue. The company aims to build a universal verification layer for cutting-edge AI labs, critical infrastructure, and national-level applications.

According to reports, Attestable uses Zero-Knowledge Proof technology to shift trust in AI systems from data centers to a mathematical verification mechanism. This technology can prove that an approved model, model weights, input data, and execution policy indeed produced specific outputs without revealing model parameters or user privacy data. Additionally, the model does not need to be rerun to complete the verification.

Attestable stated that its technology has currently achieved verification inference of the Meta Muse Glimmer 30B model on a single NVIDIA H100 GPU, achieving a speed of 85 tokens per second. The generated proof files are small, possess post-quantum secure properties, and can complete verification quickly.

In response, Ethereum co-founder Vitalik Buterin expressed in a post that this achievement indicates that the performance overhead of large-scale language model zero-knowledge proofs has reached close to single-digit levels. He mentioned that the next challenge is to further reduce the performance costs of technologies such as Fully Homomorphic Encryption (FHE).