According to @PANewsCN, decentralized trading protocol Bunni has formally shut down following a recent security exploit and a funding shortfall. source: https://x.com/bunni_xyz/status/1981160279871558114 https://twitter.com/PANewsCN/status/1981190491577471240 Bunni stated that relaunching would require six to seven figure spending on audits and monitoring plus long development and business cycles, leading the team to close the protocol. source: https://x.com/bunni_xyz/status/1981160279871558114 Users can still withdraw assets from the official website. source: https://x.com/bunni_xyz/status/1981160279871558114 The remaining treasury will be distributed based on a snapshot to BUNNI, LIT, and veBUNNI holders, excluding team members, with details to be released after legal procedures conclude. source: https://x.com/bunni_xyz/status/1981160279871558114 https://twitter.com/PANewsCN/status/1981190491577471240 Bunni v2 contracts have switched from BUSL to an MIT open-source license, and the team said it will continue assisting law enforcement to recover stolen funds. source: https://x.com/bunni_xyz/status/1981160279871558114

According to DeepLearning.AI, Microsoft has released three open-weight reasoning models—Phi-4-reasoning, Phi-4-reasoning-plus, and Phi-4-mini-reasoning—along with comprehensive training methodologies. These models outperform competitors on mathematical reasoning tasks and are freely available under the MIT license (Source: DeepLearning.AI, May 16, 2025). For crypto traders and AI-focused investors, this move could accelerate AI development in decentralized applications, potentially reducing costs for blockchain projects integrating advanced reasoning models. The open-source nature may foster innovation in on-chain analytics and DeFi automation, impacting the competitive landscape for AI-powered crypto protocols.