MetaArena has completed a new round of strategic financing to accelerate the development of its trusted AI engine.

Source: Cointime Original Title: MetaArena Secures New Strategic Funding to Accelerate Trusted AI Engine Development Original Link: MetaArena is a trusted execution infrastructure designed for AI agents and complex interactive systems. The project aims to accelerate the application of AI in areas such as gaming and finance, ensuring that intelligent behavior is verifiable, auditable, and scalable through zero-knowledge proofs, providing a truly reliable AI experience for blockchain games.

MetaArena is one of the few infrastructure projects focused on credible AI execution in blockchain gaming, which has recently gained widespread attention in the market and successfully completed a new round of strategic financing. This round of financing involves several well-known institutions, including IBC Group, Central Research, SEI Foundation, Sky Ventures, and Stratified Capital.

This round of financing not only reflects the strong recognition of the capital market for MetaArena's vision but also further validates its potential value and strategic position in the wave of on-chain intelligent interaction upgrades.

MetaArena: Building Trusted AI Agents

MetaArena is a trusted execution infrastructure centered on zero-knowledge proof (ZKP) mechanisms, providing verifiable computing services for AI games and intelligent interactive scenarios that require trusted execution.

MetaArena consists of an off-chain computing network made up of distributed nodes and an on-chain validation engine deployed across multiple chains. When a trusted execution task is generated, MetaArena distributes AI behavior requests to off-chain nodes for execution, generating zero-knowledge proofs (ZKP), which are then verified on-chain. This mechanism ensures that the input data, inference process, and execution results are all authentic, reliable, and tamper-proof. MetaArena has been validated in the Web3 gaming field, enabling AI Agent-driven blockchain games to operate efficiently, securely, and auditable without relying on centralized servers.

In a recent upgrade, MetaArena launched a new Trusted Execution Stack. Through its two core capabilities, zkTrace and zkAction, it ensures the consistency of prompt input (Proof of Prompt) and the credibility of reasoning behavior (Proof of Inference), thereby proving the authenticity and confidentiality of AI Agent prompts and reasoning paths.

It is worth noting that while many existing solutions attempt to provide a trusted environment for AI Agents, MetaArena is one of the few projects that achieves trusted execution entirely through zero-knowledge cryptography without the need for specialized hardware.

zkTrace: Proof of Trustworthiness for Prompt Input

In traditional AI Agent models, there is a long-standing unresolved core issue: how to ensure the credibility of prompts?

This includes but is not limited to:

• Was the prompt tampered with before or during execution?
• Does the model really infer according to the expected prompts?
• Is there a risk of sensitive content in the prompt being leaked?

MetaArena provides verifiable and trustworthy execution capabilities for prompts through the zkTrace module at the computation layer, ensuring the correctness, consistency, and privacy of prompts throughout their lifecycle without exposing the original content externally. This is a key foundational component for building trustless AI Agents and decentralized application logic.

zkTrace is provided in a developer-friendly SDK form. Its underlying design relies on strong cryptographic mechanisms and ZK primitives, including Pedersen commitments, Poseidon, and zkSNARKs (Plonk), and is closely integrated with the system prompt initialization process.

During the system initialization period, the prompts are input into the off-chain computation network to generate cryptographic commitments and build the corresponding ZKP. These ZKPs can be referenced by any user or third-party verifier, confirming the authenticity and untampered status of the prompts by comparing them with the on-chain prompt commitments. If the prompts used in execution do not match the audited commitments, the verification fails immediately, ensuring transparency and trusted execution without exposing plaintext.

In practice, AI Agent developers or AI prompt application developers can use zkTrace to create and define system prompts, ensuring that the model strictly executes tasks according to established policies and constraints. Once the system prompts are initialized and loaded into the model, zkTrace automatically generates commitment and proof documents, submitting them to the on-chain verification engine. This process records the complete lifecycle of the prompts from input to usage, ensuring that the proofs are traceable and tamper-proof.

For end users interacting with AI Agents, they can always access the prompt word commitments and proofs corresponding to the currently executing model, and verify the authenticity of the prompt word usage:

• Does it still match the expectations set by the developers?
• Was it replaced or injected with malicious content during execution?

zkTrace ensures that trust in prompts no longer relies on centralized custody or endorsement from a single service provider, but is instead established through cryptographic proof, creating a verifiable, auditable, and non-repudiable trust foundation for system inputs.

zkTrace Interaction Example

zkTrace establishes a reliable interaction mechanism between AI Agents, off-chain computation networks, DApps, and smart contracts, ensuring the integrity and consistency of prompts, and providing verifiable trust guarantees for AI model behavior.

When AI Agent developers define and submit system prompts through zkTrace, the prompts are encrypted off-chain, generating a commitment, while the Agent is initialized and bound to the corresponding verification circuit, ensuring that the prompts possess immutable properties throughout the system. The AI Agent also registers the necessary verification keys with the MetaArena off-chain computing network for subsequent verification calls.

When a DApp initiates a message or interaction request, the AI Agent reads the request and delegates the execution task to off-chain computing nodes. During execution, the use of prompts and logic is verified through zero-knowledge proofs, and the behavioral path is recorded to generate verifiable proof documents. The proof results are then returned to the smart contract or DApp, confirming at the contract level that the operation strictly originates from the committed prompts.

The on-chain verification engine of MetaArena is responsible for matching ZKP with commitments, ensuring consistency between the input content and the execution behavior. If the prompt is replaced or the execution strategy deviates, the verification fails immediately, effectively preventing potential anomalous operation chains. This mechanism ensures that the execution of the AI Agent is completely consistent with the initial settings, providing a transparent, auditable, and trustworthy foundation for various Web3 use cases.

By collaborating with smart contracts and other on-chain objects, MetaArena enables AI Agents to perform with public verifiability, providing high security and structured trust for various Web3 use cases.

From a capability perspective, zkTrace enables AI Agents to have:

• Data Privacy: The content of the prompt can be verified without disclosure, avoiding the leakage of sensitive information.
• Credibility and Transparency: Zero-knowledge proofs ensure that the model's behavior is not maliciously tampered with.
• Distributed validation capability: Any user or third party can verify execution consistency, avoiding reliance on centralized entities.

Based on the trusted input advantages of zkTrace, the capability can naturally extend to inference proofs (achieved through zkAction), to validate the credibility of the AI Agent's inference paths and results, ensuring that the output strictly derives from legitimate input reasoning.

Overall, zkTrace is particularly suitable for mission-critical scenarios, such as finance-sensitive, tightly constrained, or highly compliant decision-making tasks, providing a highly secure and transparent operational foundation for the next generation of trustless AI Agents.

Trustworthy Framework for AI Agent Game Engine

MetaArena has made its debut in the blockchain gaming field by launching an AI game engine component, which constrains and audits in-game agent operations through a zero-knowledge proof mechanism. Game agents can participate in on-chain battles via smart contracts, with their actions verified by zkTrace/zkAction to ensure fairness, authenticity, and traceability.

In this game engine system, developers can continue to use native engines such as Unity, Cocos Creator, and Unreal without changing their existing workflows to migrate games to a trusted blockchain environment. Through the SDK interface, developers can access the decentralized state layer of MetaArena, manage key on-chain states including player inputs, state changes, and turn transitions, and verify them in real-time using zero-knowledge proofs.

All generated content and task feedback can be processed by multiple AI Agents (such as content generation Agent, battle Agent, testing Agent), achieving automated validation and dynamic game experience optimization.

All data generated during the game process—including command inputs, state transitions, behavior logs, and content generation results—are transmitted to MetaArena's off-chain computing network for processing and integrated into a verifiable proof structure through the ZK game SDK. Using ZK circuits (such as ZK Shuffle and operation validity circuits), it ensures randomness, fairness, and rule consistency. The on-chain verification engine publicly confirms the authenticity of each action through zero-knowledge verification, ensuring that the game execution process is tamper-proof and fully transparent.

In the computing and storage layer, MetaArena combines resource optimization components to provide high-performance support for multi-agent scenarios (AIGC, QA testing agents, data insight agents, etc.), ensuring execution efficiency and response stability under high throughput interactions.

Ultimately, this infrastructure not only provides developers with efficient computing resources but also ensures that every game operation is verifiable, auditable, and accountable through decentralized validation + smart behavior auditing, establishing a fair and trustworthy on-chain AI gaming ecosystem that effectively prevents cheating, tampering, and opaque execution.

Excellent Security

When building trustworthy AI agents, TEE (Trusted Execution Environment) solutions are widely adopted due to their hardware-isolated environments, providing a certain level of data privacy protection and verifiable execution. Although TEE is a mainstream privacy solution for cross-domain verification, there are limitations in its application for building trustworthy AI agents.

In fact, TEE solutions typically rely on trusted environments and key management services provided by hardware vendors such as Intel SGX and ARM TrustZone. This centralized trust mechanism makes system security highly dependent on specific vendors, introducing centralized risks. Intel SGX has been exposed to multiple vulnerabilities that directly threaten the trust foundation. Additionally, although TEE provides an isolated execution environment, its data privacy protection remains limited. For example, data may be intercepted when transmitted to the TEE environment, and external attackers may access sensitive information through interactive interfaces. TEE designs are primarily targeted at predefined computing tasks, lacking dynamic adaptability, while AI Agents often handle various tasks and complex contexts, which is difficult to support for rigid architectures.

In contrast, MetaArena's zero-knowledge trusted execution solution is decentralized and does not rely on any centralized entities. Its security comes from a large-scale off-chain distributed computing network. This provides lightweight advantages, excellent scalability, and flexibility compared to TEE, allowing it to efficiently adapt to diverse AI Agent applications. MetaArena seamlessly supports popular models like ChatGPT and DeepSeek. Notably, MetaArena's solution is entirely based on ZK cryptography, making it stand out in the field of trusted AI Agents.

Overall, although AI technology is evolving rapidly, the large-scale adoption of fully autonomous AI Agents still faces challenges in terms of safety, ethics, and practicality. The balance of automation and human oversight with semi-autonomous AI Agents remains mainstream. This means that before large-scale adoption, AI Agents need to make progress in trust and privacy. MetaArena is accelerating this process with its fully ZK-based cryptographic solutions, laying a solid foundation for the next stage of AI Agent development. The new round of financing further establishes its position as a leading trusted AI engine infrastructure.