ZetaChain has announced a major strategic shift, abandoning its cross-chain blockchain infrastructure business to focus entirely on artificial intelligence.
As part of the transition, the company unveiled Anuma, a consumer-facing AI platform designed to give users a portable, encrypted memory layer that works across multiple leading AI models and assistants.
The move addresses a growing challenge in the AI industry: users often spend months building context, preferences, and workflows within one AI system only to lose that personalization when switching to a different model.
Anuma aims to solve that problem by creating a persistent user-owned memory vault that can travel across platforms including OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, xAI’s Grok, and DeepSeek.
According to ZetaChain, the application attracted more than 60,000 users during its first month following a public launch in April.
To support the new strategy, ZetaChain will gradually phase out its existing interoperability infrastructure and discontinue support for multiple blockchain networks, including Bitcoin, Solana, Base, Arbitrum, BNB Smart Chain, Polygon, Avalanche, Sui, and TON. Asset withdrawals will remain available during a structured migration period.
The company said its native ZETA token will continue to serve as the foundation of the ecosystem, powering identity management, permissions, payments, agent interactions, and decentralized storage. The platform also plans to leverage the x402 payment protocol to support machine-to-machine microtransactions and AI agent commerce.
The announcement reflects a broader industry shift toward AI infrastructure and agentic systems. Recent developments include Coinbase’s rollout of Base MCP for AI-powered blockchain interactions, the launch of x402 batch settlement for autonomous AI payments, and growing investment in AI-native financial rails from companies including Coinbase, Circle, AWS, and BNB Chain. As competition intensifies around AI agents and personalized assistants, ZetaChain is betting that user-owned memory, identity, and context will become critical infrastructure for the next generation of AI applications.
