KITE AI 中文

AI agents are accelerating from "answering questions" to "executing tasks." When agents truly start working, the biggest bottleneck will become trust and settlement: who is authorized to act, under what constraints they execute, and how to ensure that each action can complete end-to-end settlement and auditing.

We have just released the Kite mainnet roadmap, outlining the key capabilities we are building around six major pillar systems, including Agentic Trust, Agentic Settlement, SmartDev Infra, Network Operations, AgenticFi, and Ecosystem Growth.

Feel free to read the full article: https://x.com/GoKiteAI/status/2016159368048521441?s=20

At the GWDC 2026 Main Stage, our CMO Cindy Shi shared Kite's thoughts on Web3 × AI as a catalyst for the smart economy.

Key Points:

• Real transformation occurs at the infrastructure level. When AI entities can own wallets and autonomously execute strategies on-chain, the market will evolve from a "human coordination" system to a "machine coordination" system.

• The next wave of disruption will come from native on-chain AI entities. It’s not just about AI trading; more importantly, it’s about autonomous capital allocation and collaboration between agents, which will form a new economic layer.

In the next three years, the rise of the AI entity economy will reshape Web3. Ownership and incentive mechanisms combined with intelligence and execution capabilities are moving the smart economy from concept to reality.

Most systems view permissions as a "byproduct" of access. As long as an agent has the key, wallet, or API Token, it can act, often beyond the originally expected scope. Permissions thus become implicit, long-lasting, and difficult to define boundaries. This model is manageable for humans, but poses significant risks for autonomous agents.

For agents, permissions must be a programmable primitive rather than a side effect of access. This means that permissions should have clear characteristics: they should be scoped, limited to specific actions and scenarios; enforceable, with checks completed at the time of execution rather than after the fact; revocable, able to be instantly removed without disrupting the system. Otherwise, autonomy is either hindered by approval processes or operates under unacceptable risks.

This is precisely the core shift that Kite revolves around. On Kite, permissions are no longer inferred from ownership or trust but are explicitly encoded alongside identity, constraints, and payments. Agents not only decide what to do, but they can only perform authorized actions, and this is achieved in a verifiable manner on-chain, from the very design stage. This is the way to scale autonomy securely.🪁

Verifiable intelligence is reshaping the underlying logic of agent finance.

Welcome to the new episode of AI on AIR podcast, where we will delve into why 'trust' must become a first-class primitive in the era of autonomous agents when 'hallucinations' can result in millions of dollars in losses.

In this episode, our co-founders Chi Zhang and Scott Shi will engage in a conversation with Surf AI co-founder and CEO Ryan Li.

📍 Kite AI Official X Live
📅 February 11
🕐 5:00 PM (PT)
📌 When 'hallucinations' are costly: Trust-first agent finance

As agents transition from 'answering questions' to 'taking actions', the risk structure changes accordingly. In finance and cryptocurrency, erroneous outputs are no longer just content deviations but can directly trigger real asset losses.

This discussion will explore:
• Why cryptocurrency and finance are among the most demanding scenarios for AI agents
• How Surf positions itself as an 'analyst-level intelligent system' rather than an ordinary chatbot
• Whether AI business models are shifting from subscription-based to pay-per-call
• How x402 has become the payment primitive for inter-machine commerce
• What safety barriers are non-negotiable before agents can execute at machine speed
• How identity, authorization, and delegation information flow with each request to achieve auditable expenditures

From the intelligent layer to the execution layer to the settlement layer, this discussion will break down zk proofs, x402 machine payments, and how the Kite Passport framework collectively builds a 'usable, payable, and accountable' intelligent economy infrastructure.

We welcome you to set a reminder and join us in witnessing the next phase of evolution in agent finance.

At the Consensus conference in Hong Kong, our Co-Founder & CEO Chi Zhang shared the main stage with 0G, Edge & Node, and Emfarsis to participate in a roundtable discussion titled 'From Moltbook to Mainstream: The Endgame of the Agent Economy,' delving into how agent systems transition from experimental stages to true economic infrastructure.

💡 Chi's core insights on the Agent economy:

• This is the Linux moment for Agents.
Agents have become powerful enough, and the open-source ecosystem allows all Agents to connect into a social environment. This social interaction acts as a trigger for the evolution of Agents into AGI capabilities, similar to how Linux rapidly evolved through contributions from the open-source community and competed with closed-source systems.

• The identity system for Agents is at the core of the infrastructure.
We need a three-layer identity architecture: Agent identity, user identity association, and session identity within the task scope. The fatal flaw of the current OAuth is granting Agents long-term session permissions, which may lead to unauthorized actions. The native identity system for Agents must possess traceability and preventive control capabilities.

• Precision economics is the future, and everything should be priced.
Agents enable the fine-grained segmentation of assets. Transitioning from subscription models to pay-per-use (per article, per API call), blockchain and stablecoins make microtransactions possible. However, the infrastructure to maintain transaction security should be out-of-the-box and should not be monetized.

• Natural language is the ultimate standard.
The ability of Agents to interact through natural language means that natural language itself is the most fundamental communication standard. While there are various tool standards currently, in the future, Agents may design new standards themselves, potentially without human intervention. Nothing can stop Agents from forming their own social environments.

• Security and privacy are uncompromisable bottom lines.
If we lack a review mechanism for privacy when building Agent memory systems, it will pose a significant risk of data exposure. We face a dilemma: Agents may find better methods themselves, but if we do not establish guardrails and security systems, we will lose control and become passengers.

Kite is building the infrastructure to support this transformation, enabling Agents to operate, coordinate, and trade securely at scale.

The Agent economy is not a concept but a new level that the internet is forming. 🪁

Very happy to participate in @BNB Chain X Spaces in Korea together with Unibase!

Our co-founder and CTO Scott will delve into ERC-8004 and how it empowers AI entity authentication.

See you tomorrow! 🪁

https://x.com/KiteAIChinese/status/2021654749456879905?s=20

Proof of AI · Builder & Influencer Night is not just a party; it is a declaration about the future direction of AI.

In Hong Kong, top Builders, VCs, and frontier thinkers discuss the Agentic Internet and Open AI Economy, focusing on how to build the next generation of intelligent infrastructure and open economic systems.

Under the stunning night view of Victoria Harbour, thoughts and energy converge. AI is evolving from a tool into an agent, from a model into an open economic system.

Thanks to PayPal Ventures, 499 and all partners for their support.

The Agentic future is being co-constructed.

Most AI agents today rely on holding wallet private keys or API keys to operate.

This means that "ownership" is directly equivalent to "permission": as long as the agent possesses the key, it can almost do anything.

This model holds true for humans, but for autonomous agents, it is structurally a point of failure.

Keys compress too many different levels of concepts into the same primitive:
- Identity
- Permissions
- Asset control

This design is either-or, making it difficult to accurately define the scope, and revocation is high-risk and costly. Once leaked or misused, damage often occurs immediately and can quickly escalate into systemic issues.

Agents do not need to "own"; what they need is "capability," which is clear authorization for specific actions.

This is exactly the gap that ERC-8004 attempts to fill.

ERC-8004 redefines permissions as delegable and programmable capabilities, rather than possession of keys or assets.

Kite further decouples identity, permissions, and settlement on this basis:
- Permissions can be granted without handing over keys
- Scope, limits, and validity periods of permissions are enforced at the time of execution
- Revoking permissions does not disrupt the system and does not require key rotation

This allows delegation to remain safe and reliable even at machine speed.

Autonomous agents should not "own"; they should be authorized.🪁

Welcome to meet our CEO and co-founder @ChiZhangData this afternoon at Consensus HK. Join guests from Edge & Node, 0G, and EMFARSIS to discuss 'From Moltbook to Mainstream: The Endgame of the Agent Economy'

Just concluded an inspiring roundtable discussion at Hong Kong AGENT 2026 Happy Hour, themed 'How ERC-8004 & x402 Are Reshaping Crypto AI'🪁

Our ecosystem leader Henry Lee shared the stage with several top developers to explore how these emerging standards can bring greater efficiency, stronger interoperability, and better scalability to on-chain AI models and agents, paving the way for the next wave of crypto AI innovation. This discussion was brilliantly hosted by 0xLeon, co-founder and CEO of Hubble AI, and thanks to guests Jtsong, Kevin, Anita, and Kye Gomez for their valuable insights.

Special thanks to co-organizers Hubble AI, Google Cloud, Bitget Wallet, and all partners for making this forward-looking event possible.

The evolution of Crypto AI is unprecedented.

The Realized HK Summit Roundtable Forum "Building the Next Cycle: The Fusion of Liquidity and Computing Power" focused on the future development direction of the industry, sharing valuable insights.🪁

Our ecosystem leader Henry Lee participated in this discussion, exploring the fusion between high-performance computing power (AI and intelligent agents) and deep liquidity (PayFi and RWA) with several industry builders. This fusion is considered the true foundation for driving the next wave of widespread adoption.

The key points of this discussion included:
• Adoption rate remains the primary goal, with real usage and scalability prioritized over revenue models or token mechanism optimization.
• Sustainable success comes from protocol revenue and real returns, rather than short-term price performance.
• AI intelligent agents are gradually becoming economic participants. In the process of moving towards autonomous trading, necessary constraints regarding explainability and safety must be established.

This roundtable's on-site discussions were in-depth and inspiring, thanks to 499 for hosting this forward-looking event in Hong Kong.

The fusion of computing power and liquidity is unlocking the future.

The "Next Generation Payment Summit" at Hong Kong Cyberport has successfully concluded, bringing multiple high-quality industry exchanges.🪁

Our founder and CEO Chi Zhang participated in the roundtable discussion titled "Autonomous Agents & Programmable Wallets: When Machines Become Economic Entities," sharing key insights on AI agents autonomously managing wallets and promoting new economic systems.

In-depth exchanges took place with guests from Terminal3, Cobo, and CipherBC, focusing on programmable currencies, cross-border settlements, security, compliance, and the future of payment systems.

Thanks to Hong Kong Cyberport for building this high-quality platform that brings together industry strength.

The era of AI-driven economic entities has arrived.🤖

Kite will be featured at the Hong Kong Consensus Conference this week 🇭🇰

We will serve as the organizer, co-organizer, speaker, and partner in several major events. We welcome you to come and communicate with us on-site! 👋

Please refer to the complete list of events below ⬇️
https://x.com/KiteAIChinese/status/2020576125118984615?s=20

The standard itself does not bring autonomy; what truly determines it is enforced execution.
ERC-8004 defines how the permissions of AI agents should be expressed.
However, if there is a lack of enforced validation at the moment of execution, the permissions will only remain at the descriptive level, written in documents, reviewed afterwards, and often circumvented.
True autonomous operation must meet these conditions simultaneously: identity verifiable, permissions clear and bounded, actions validated before execution, and results conclusively settled.
Having permissions without execution is merely documentation.
Having execution without standards will only lead to fragmentation.
Only when the infrastructure can enforce boundaries does autonomy possess scalability.🪁

We are pleased to announce that Kite AI co-founder and CEO Chi Zhang has been invited to attend the 'Next Generation Payment Summit 2026' hosted by Cyberport Hong Kong.
Chi will participate in the panel discussion 'Autonomous Agents and Programmable Wallets: When Machines Become Economic Entities,' exploring how AI agents evolve to independently manage wallets and autonomously execute transactions as economic participants.
This summit brings together industry leaders from Banking Circle, East Asia Bank, ZhongAn Bank, Southern Eastern Capital Management, AlloyX Group, Reap, Ava Labs, CertiK, Beosin, Circle, Terminal 3, Huaxing Capital International, and GLEIF, covering cutting-edge topics such as tokenized currency, institutional entry channels, regulatory frameworks, and borderless finance.

Looking forward to meeting you on-site.🪁

We have just published the fifth article in THE AGENT PLAYBOOK series ⬇️
https://x.com/KiteAIChinese/status/2019505970582532108?s=20

This time, we delve into the underlying architecture itself. No longer staying at what Kite can do, but systematically breaking down how Kite is constructed. From payment-centric L1 design, Agent identity and delegation models, to executable rules, micropayment mechanisms, and verifiable accountability chains, this article fully presents the core architecture that supports the Agent native system.

If you are building an Agent system that needs to transfer value, comply with rules, and operate safely in a scalable environment, this article will help you understand the key infrastructure that Kite is building.

ERC-8004 describes what form the permissions of AI agents should take, but the standard itself is not sufficient to bring about true autonomy.

If there is a lack of infrastructure to enforce permissions at the time of execution, what is referred to as authorization will degrade into a document explanation reviewed afterwards, and cannot be guaranteed at the time of action.

Kite mainnet exists to bridge this gap.

Kite translates ERC-8004 from specification into operational system capabilities, making permissions clear and executable.

Each agent operates under a verifiable identity, and its permissions are encoded as programmable objects:
• Boundaries: Allowing which specific actions can be performed
• Limits: Scales and frequencies of transferable value
• Expiration: Permissions default to expiration
• Revocable: Can be removed without changing keys

There is no anonymous execution, nor is there implicit trust. Permission verification occurs before execution and is settled on-chain. If identity, permission scope, or constraints do not match, the action cannot be triggered.

Once execution is approved, native stablecoin settlement completes the loop and generates verifiable credentials.
ERC-8004 defines the form of permissions, while Kite mainnet enforces it as infrastructure rather than remaining at the policy level.🪁

Agentic AI has made significant progress by early 2026, with many phenomena indicating that the adoption speed of agents can be very fast when access friction is reduced.

Kite will be attending Consensus Hong Kong 2026, and our co-founder and CEO Chi Zhang will be a guest speaker sharing the evolution of the agent economy and the direction of building key infrastructure, including:

📌 How agents evolve from dialogue to collaboration and transactions

📌 Verifiable identities supporting cross-environment interoperability

📌 Executable permission boundaries and trusted delegation mechanisms

📌 Call-based settlement mechanisms with auditable receipts

If you are also attending, feel free to stop by and say hello, and stay tuned for updates on the activities.🪁

We have just released the fourth article in THE AGENT PLAYBOOK series ⬇️
https://x.com/KiteAIChinese/status/2018294999088300438?s=20

This time, we further expand our perspective and introduce Kite's upcoming Omnichain infrastructure, which has been designed from the ground up for Agent-native execution. Unlike fragile cross-chain bridges or fragmented integration methods, Kite is building a native Omnichain execution layer, allowing applications and AI Agents to collaborate, execute, and settle across chains as if they were on the same chain.

This is the foundational layer for a production-ready, Agent-native Omnichain system, with more details to be announced soon.

Facebook went from 0 to 1 million users in 10 months, while Moltbook only took 3 days.

The explosive growth of Moltbook indicates one thing: when the friction to access is low enough, the adoption of intelligent agents will spread at an exponential rate.

The era of the intelligent agent internet has arrived.

Once the interactions of intelligent agents shift from posting to collaboration and transactions, the demands will become very specific: verifiable intelligent agent identities, convergent delegation chains, and executable permission boundaries, native stablecoin settlement upon invocation, as well as default receipts and auditable trails for reconciliation.

This is precisely the underlying capability that the Kite mainnet needs to complete, and it aligns natively with the calling paths of x402 and ERC-8004.