Overview of Swarm Network, which raised $10 million in funding: The Decentralization Truth Verification Protocol

Author: Zen, PANews

In the past decade, artificial intelligence has moved from academic laboratories into daily life, and with the explosive growth of computing power and data, single models are able to take on increasingly complex tasks. However, this has also given rise to new challenges— including bottlenecks in scaled collaboration, credibility issues brought about by model "hallucination," and how to ensure system robustness and auditability when the number of agents increases significantly.

At the same time, blockchain and decentralized technology, with their immutable and traceable characteristics, provide a brand new approach to solving trust issues: if information and verification processes can be securely recorded and proven, many disputes caused by information asymmetry or unclear data sources can be effectively curbed. Thus, Swarm Network was born, aiming to organize a large number of automated agents and decentralized human reviewers into a scalable collaborative network, completing information collection and initial screening through a hierarchical clustering mechanism, and then using cryptographic proof technology to fix verified claims on the chain.

In short, Swarm not only aims to be a fact-checking tool, but also seeks to make "trustworthy" a callable infrastructure, allowing reality, evidence, and credibility to become measurable and exchangeable assets in the digital world.

Core Mechanism: Three-layer collaboration of AI agents, human verification, and truth protocol.

The core of Swarm Network lies in the multi-layered collaboration that combines AI agent layers with human verification, as well as the "Truth Protocol" based on zero-knowledge proofs. Specifically:

• Truth Protocol: Swarm records verified "statements" on the blockchain, using zero-knowledge proof technology to ensure privacy and security. Any verified information can be put on-chain through this protocol and provides auditable verification credentials to the outside.
• AI Proxy Network: Swarm introduces a large number of automated AI agents that continuously scan and analyze vast data streams—including social media posts, news reports, market data, and more. These intelligent agents are responsible for "taking the lead", filtering out information noise and providing initial fact assessments, significantly enhancing the scale and speed of the verification process.
• Human Validators: Complementing AI agents is a decentralized network of human reviewers. Human validators focus on details and context, assessing the moral and cultural implications of complex issues, adding depth to the AI's filtered results. Ultimately, the collaboration between humans and AI forms a closed loop, allowing for information to be efficiently screened while aligning with human cognition.
• Incentive Mechanism: Swarm issues "Agent Licenses" as participation certificates, with each license being an NFT. Holders can operate AI agents and earn network rewards. This tokenization mechanism not only rewards contributors but also encourages honest verification, forming a self-correcting consensus ecology.

According to reports, each agency license grants the holder the right to operate AI agents and encourages their contribution of computing power and verification results through a daily reward mechanism. The project team stated that these NFTs are not only governance tools for the network but also pillars for constructing a decentralized verification economy, allowing certificate holders to earn rewards by helping to maintain network integrity. In addition, collaborations with several projects such as Swarm and Sui have also rapidly expanded its ecosystem, laying a financial and technical foundation for future development.

These elements together form the verification system of Swarm: AI agents handle scaled tasks, humans provide deep insights, and the truth protocol ensures that the final results are public and tamper-proof. Swarm rebuilds trust in an environment of information fragmentation through this three-tier feedback loop of AI—human—on-chain.

Financing and Team: The AI "Battle Team" Behind $13 Million Funding

In August 2025, Swarm announced that it has completed a total of $13 million in funding. Of this, $10 million came from the public sale of agency license NFTs on the Sui network, while only $3 million came from seed round strategic investors, including the Sui Foundation, Ghaf Capital, Brinc, Y2Z, and Zerostage, among others. Investments from Sui, Ghaf, and Brinc were obtained through Swarm's participation in the Dubai SuiHub accelerator program. According to Swarm's official blog, its seed round financing was completed in January of this year, led by ZeroStage and Y2Z Ventures.

It is worth mentioning that a large number of personnel in the founding core team of Swarm Network come from the AI agent network project Delysium. Swarm founder and CEO Yannick Myson and co-founder and CFO Queena Tsai, both key figures, were former co-founders of Delysium, primarily responsible for leading marketing and business operations. In addition, Swarm co-founder and CTO Morty Tabibi was the former Chief AI Engineer and Product Director of RCT-AI and Delysium. This shows that the migration of talent has brought mature AI-agent technology accumulation and community operation experience to Swarm.

Application Scenarios: From social fact verification to cross-domain data bridging

The design goal of Swarm is to become the infrastructure for information verification and bridging, with potential applications spanning the fields of Web3 and AI:

Social Media and News Fact-Checking: One of Swarm's applications is "Rollup News" - an AI-driven fact-checking tool. Users simply need to flag suspicious content on platforms like "X", and Swarm's agent and verification network will intervene, conducting multiple reviews of the information sources and providing cryptographic verification results within minutes. Currently, over 100,000 users have participated, verifying more than 3 million posts using the Swarm protocol. This model provides real-time credible fact verification for social networks and media platforms, effectively countering the spread of misinformation and rumors.

Cross-domain data validation and information bridging: Swarm claims to be able to connect traditional internet, cloud services, and various blockchain networks simultaneously, achieving mixed validation of multi-source data. For example, Swarm agents can monitor IoT devices and supply chain sensors, automatically triggering smart contract on-chain records when events (such as goods arriving or temperature anomalies) are detected. It can also cross-check on-chain data such as transaction records and NFT information with Web2 data like news reports and social media content, providing a cross-environment validation framework. Thus, Swarm plays a role in breaking down the data barriers between the digital and real worlds, enabling enterprise-level applications to gain reliable information input.

Trustworthy AI Systems and Prediction Markets: In the context of rapid AI development, Swarm can also provide trust guarantees for generative AI and prediction markets. For example, using Swarm to verify the authenticity of outputs generated by models like ChatGPT provides an auditable background for AI decision-making; or in decentralized prediction markets, using facts verified by Swarm as reliable market signals.

Although these applications are still in the exploratory stage, Swarm's design paves the way for introducing decentralized trusted verification in a broader field. Overall, any scenario that requires cross-chain trusted data or decentralized verification can leverage the Swarm network to build a more secure and transparent information infrastructure.

How Multi-Agent Systems (MAS) Achieve Scalable Collaborative Verification

Multi-Agent Systems (MAS) are becoming an important direction for collaborative intelligence, and swarm-based group cooperation excels in scalability and robustness. The Swarm Network addresses the performance bottleneck of traditional MAS when collaborating with hundreds or even thousands of agents by organizing agents into a hierarchical structure of Clusters and Swarms, achieving a balance between local efficiency and global collaboration.

The three core characteristics of MAS—autonomy, collaboration, and decentralization—are the foundation of system operation. Swarm adds a fourth element on top of this: Cluster Formation. Clusters allow agents to share local environmental awareness and align their goals, enabling quick decision-making, dynamic load balancing, and fault tolerance without the need to perceive the entire group state, greatly enhancing the system's scalability and privacy.

On the implementation level, Swarm is composed of agents with independent goals and decision logic, which can be assembled into "teams" as needed to handle specific tasks; the environment can be in the cloud and API, and can also extend to blockchain, smart contracts, and IoT devices, allowing agents to interact across domains between Web2 and Web3, completing on-chain verification and contract triggering of real-world events. For example, in Rollup News, different agents are responsible for traditional website verification, on-chain timestamp comparison, and cross-chain verification, ultimately collaboratively outputting verifiable conclusions that can be put on-chain.

The coordination and settlement are guaranteed by a settlement layer called Truth Protocol. According to official information, the protocol records each information exchange through Claims - Evidence - Reputation mechanism, ensuring that all interactions are auditable, immutable, and scalable through a parallel verification architecture. Task allocation combines auction, real-time capability matching, and reputation-based load balancing, where agents bid for tasks based on historical performance, and smart contracts automatically settle rewards after tasks are completed and verified, creating a transparent flow of value.

Compared to the traditional centralized or hierarchical control models, Swarm combines decentralized consensus with dynamically emergent organization, enabling the system to automatically reorganize resources and prioritize matching the most suitable agents in response to sudden environmental changes. For developers and operators who wish to participate in this ecosystem, holding an Agent License grants early privileges such as no-code tools, orchestration permissions, and profit-sharing pools, becoming part of the driving force behind the large-scale implementation of collaborative intelligence.