Fhenix, a prominent layer 2 platform that uses FHE cryptography to power Ethereum developers, has started a new partnership. According to the platform, it is joining forces with Privasea (a top player in FHEML-centric AI+DePIN computing) to promote the secure development of AI apps using FHE technology. The company shared an extensive blog post revealing the details of the partnership.
Fhenix and Privasea are working together to improve the secure development of AI apps
The company noted that a large amount of data comes online regularly, highlighting the need for data privacy technologies. In addition, the demand to secure digital sovereignty has increased due to the advent of artificial intelligence and large language models. The respective models reportedly use massive amounts of data for training.
Fully homomorphic encryption works as the best encryption solution in artificial intelligence. It allows instant calculations when dealing with encrypted data. This allows LLMs to process the massive amounts of data required for blind operations, without the need for decryption. This paves the way for a secure relationship between the AI and the consumers.
The initiative leverages the companies’ expertise and increases interoperability between them
The partnership between Privasea and Fhenix aims to benefit from the exclusive expertise and talents of both companies. In this way, it will reportedly promote development by joining both blockchain and AI technologies. Efforts in this area take into account the ongoing development of the FHE libraries. They also include providing interoperability between the two companies.
The partnership also focuses on launching hardware acceleration initiatives. Both platforms will work together to expand Zama’s TFHE-rs library. It is a core infrastructure component of both entities. The teams from both companies will reportedly integrate Privasea’s applications on Fhenix’s L2 infrastructure. In addition, companies will delve deeper into the new mechanisms.
They will enable the integration of additional homomorphic encryption schemes in line with CKKS/BGV/BFV. They reportedly allow simultaneous SIMD processing and data packaging. Furthermore, this will better support highly accurate and large-scale computing scenarios. As a result, both entities will see a wide range of product opportunities. The partnership will use advanced cryptographic technology FHE to deliver the latest use cases such as AI.