Covenant AI announces withdrawal from the Bittensor network, questioning its governance centralization issues

ME News Report, April 10 (UTC+8), Covenant AI issued a statement saying it will withdraw from the Bittensor network and questioned its governance structure. It stated that although Bittensor claims to be decentralized externally, actual governance power remains concentrated in a few people, and related decisions lack transparent processes and community consensus. Covenant AI said that during its recent subnet operations, it encountered measures including suspension of revenue sharing, adjustments to community management permissions, and infrastructure changes, which it believes are inconsistent with decentralization principles. The team stated it will continue to promote decentralized AI training and carry out subsequent research and project deployment in other environments. Covenant AI was previously a well-known subnet project within the Bittensor ecosystem (formerly named Templar). In recent months, it completed training of the Covenant-72B model, which has 72 billion parameters, trained by over 70 independent contributors on general hardware. It is regarded as one of the largest decentralized LLM pretraining practices to date and has received public recognition from Nvidia’s CEO and the co-founder of Anthropic. The team said it will continue to promote decentralized AI training and conduct further research and development in other environments. (Source: ODaily)