KernelDAO has announced a partnership with Mira to introduce an API designed to enhance the reliability of AI systems through trustless verification. This initiative is supported by $300 million in Total Value Locked (TVL) and focuses on addressing current challenges in AI reliability.

The collaboration leverages KernelDAO's restaked economic security from its Kernel product, coupled with Mira's advanced verification technology, which reportedly offers ten times the reliability of traditional AI models. The API aims to ensure quality outputs from AI systems through a dynamic allocation of security distributed among a network of specialized node operators. These operators validate the outputs of AI models while the economic security is adjusted automatically based on their performance.

Amitej Gajjala, Co-Founder of KernelDAO, expressed enthusiasm about the partnership, stating it will help develop AI systems that assure minimal downtime for businesses and developers. The collaboration aims to tackle issues such as AI system errors and operational failures, which are critical in the current AI landscape, especially with models like DeepSeek, where accuracy can be uncertain.

Through this integration, developers will benefit from a system that includes built-in quality assurance for AI model outputs. Karan Sirdesai, CEO of Mira, emphasized that the partnership introduces real economic stakes in AI verification, providing developers with more confidence when deploying AI systems in production environments.

The API is set to become available to developers within the next year, paving the way for new applications and integrations in AI technology. This announcement follows KernelDAO’s establishment of a $40 million Ecosystem Fund, bolstered by support from various venture capital firms and investors, aimed at expanding its partner network.