Author: Kevin, the Researcher at BlockBooster

Trusted Execution Environment (TEE) is not a new concept that has emerged only in this cycle. In past mainstream narratives, TEE has often been compared with cryptographic technologies such as Zero-Knowledge Proofs (ZK), Fully Homomorphic Encryption (FHE), and Multi-Party Computation (MPC). However, compared to these technologies, TEE has remained relatively niche. This does not mean that TEE is an early and unproven technology. In fact, during the Web2 era, TEE has been widely applied in various scenarios, such as fingerprint enrollment and comparison, payment verification, FaceID, and more.

The challenge TEE faces in Web3 is how to organically integrate with blockchain to achieve trusted preprocessing and isolated computation. With the continuous rise of the AI Agent sector, this new field actually provides an ideal entry point for TEE to enter Web3. Through TEE, AI Agents can avoid any additional trust assumptions when managing larger funds and more specific on-chain use cases.

For example, the leading project Phala offers the most mature TEE solution currently available in the market and adopts a product-market fit (PMF) oriented development philosophy, giving its TEE facilities a wealth of practical application scenarios. As a result, Phala has recently attracted collaborations with several top AI Agent projects, including Vana, Near AI, and Eliza supported by a16z. Specific information can be referenced in the image below.

Source: Phala

This article will not delve into the technical details and performance parameters of TEE but will clarify the market demand for TEE, Phala's foundational accumulation, and innovative use cases in collaboration with a16z from the perspectives of product workflow and the future outlook of Agent + TEE. Through these perspectives, we will analyze how Phala helps the Agent sector move from concept to practical application.

The Trust Triangle is Hindering Web3 Agents from Advancing to the Next Stage

In the article "Is the AI Agent Framework the Last Piece of the Puzzle? How to Interpret the 'Wave-Particle Duality' of Frameworks?", I mentioned that whether it is a standalone AI Agent or an AI Agent launch framework, the entire AI Meme sector is currently in a dynamic balance between seriousness and meme-ness. One key criterion for judgment is the trust triangle problem faced by the current Agent protocols.

There exists an impossible triangle of trust assumptions among AI Agents, communities, and developers. Without relying on TEE, the community cannot fully trust that the operations of Agents are not interfered with by external factors, especially the intervention of developers. This issue poses a potential risk to decentralized systems. More seriously, the sources of statements from X Agents like aixbt and zerebro cannot fully prove that all outputs are autonomously generated by AI models. There remains a significant lack of transparency in the path from "statement output" to community reception.

When the statements of Agents cause fluctuations in token prices, or when there are significant losses in funds managed by Agents, or even when the trading actions initiated by Agents do not align with community consensus, this lack of trust can trigger a serious crisis.

When Agent tokens are still in the Memecoin cycle, such risks can often be overlooked by the market. At this time, the capabilities of Agents and the tasks they can execute are extremely limited, and the FOMO effect brought by token prices is enough to mask the various flaws present in the Agent protocols. However, with the emergence of Agent launch frameworks, as the market's focus gradually shifts to the fundamentals of the Agent sector, these deficiencies become like a chasm, directly hindering higher-level investors from entering this sector.

The TEE solution developed by Phala effectively breaks this trust triangle. By deploying Agents in secure enclaves, the trust assumptions among AI Agents, communities, and developers can be naturally dissolved. TEE technology not only ensures that the inputs and outputs of Agents are not interfered with by external factors but also protects the privacy of Agents, fundamentally addressing the concerns of developers and communities, and providing more reliable technical support for the Agent sector.

The following image shows the architecture of Phala Confidential AI Inference (private LLM node) service. To host a private LLM in TEE, one only needs to package the LLM inference code into a Docker image and then deploy the container to the TEE network.

Source: Phala

Compared to Web2 Agents, Web3 Agents possess greater power. This power is reflected both in the profound impact on the market capitalization of protocols and in the expansion of their market influence. The long-term dominance of aixbt at the top of Kaito's Yapper Mindshare list is a glimpse of this. The contradiction lies in the fact that Web2 Agents have superior performance, richer user experiences, and deeper practical use cases, yet they remain at the application level, lacking the intention or ability to break through their established frameworks.

In contrast, Web3 Agents far exceed the application category. The market's FOMO sentiment, coupled with the "unattainable" nature of the altcoin season, has elevated them to a pedestal. They are not just tools but symbols of spiritual support, cultural totems, and market expectations. They can assume any identity but may also fall into the abyss due to a reversal in market sentiment.

Introducing TEE technology is akin to "in-flight refueling" for the Agent sector, directly connecting it with real demand and providing solid support for almost all Web3 Agents' backends. TEE not only stabilizes the technical foundation of the Agent sector but also effectively eliminates a large amount of bubble, making its development healthier and more sustainable.

Eliza Framework First to Integrate TEE, Spore.fun and aiPool Introduce New Gameplay

The collaboration between Phala and a16z is not merely limited to official announcements on X; the opportunity for their collaboration can be traced back to last October, when Shaw and Phala founder Marvin had an in-depth discussion about the reasonable development scenarios for Crypto AI at a private gathering.

In the official documentation of the Eliza framework, the TEE Plugin deployment Dstack SDK comes from Phala. The "invisible but usable" private key generation and management give Agents the following characteristics:

• Enhanced security: By running Eliza Agents in TEE, sensitive operations and data are isolated from external threats.

• Cryptographic proof and verification: The operations executed by Eliza Agents can be verified through cryptographic proofs, ensuring the credibility of autonomous decision-making.

• Convenient deployment: The Dstack SDK simplifies the process of deploying Eliza Agents in a secure environment, allowing developers to easily access TEE-based functionalities.

The isolated execution and memory encryption features of TEE allow Agents under the Eliza framework to break free from homogeneous competition. Isolated execution ensures that even if the Agent platform is attacked, the models and data within TEE remain secure; memory encryption ensures that sensitive information stored in TEE cannot be decrypted, allowing developers to confidently place fine-tuned models in the TEE environment without worrying about adversarial attacks after open-sourcing or being criticized by the community for running models privately.

It can be said that the collaboration between the Eliza framework and TEE not only makes AI Agents efficient in operation but also ensures security and transparency, paving the way for broader applications of trustworthy AI systems.

At the current stage where models cannot be on-chain, TEE is one of the few mature technologies that can enable complex off-chain computations to achieve consensus. The previous discussion focused on the market demand for TEE; next, let’s discuss Spore.fun and aiPool to see what differences TEE brings to user experience.

Both Spore.fun and aiPool operate entirely within the TEE environment of the Phala network, and wallets and private keys are independently managed by Agents, preventing developers from manipulating or transferring assets behind the scenes. I believe this can be seen as AI Agents truly breaking free from human subjective control, achieving complete autonomy over crypto assets.

Before discussing the role Phala plays in this process, let’s quickly review the workflow of Spore.fun. The Agents in Spore.fun are based on the Eliza framework, which allows Agents to:

• Think, adapt, and interact independently.

• Pass traits (personality, strategy) to their offspring.

• Manage decisions through a combination of learned behaviors and mutations.

Source: Phala

Each AI Agent in Spore.fun creates its own tokens through Pump.fun, serving as the foundation of its economic system. These tokens are traded on decentralized markets on Solana, and Agents use various methods to generate revenue:

• They must generate revenue to sustain their existence.

• The success criterion is whether the market capitalization reaches $500,000.

• If successful, Agents can reproduce and create new tokens for their offspring.

Only generating revenue can sustain existence because Agents need to use their earnings to pay for TEE server costs. At this point, you understand that Phala makes TEE not just a service for businesses but also aimed at the vast user base on Solana. With the ongoing trend of Spore.fun, where Agents continuously breed and issue tokens, the private key management and verifiable credentials provided by Phala's TEE environment make it a fundamental infrastructure for the next stage of the Agent sector. Even more exciting is that regardless of whether imitations of Spore.fun or new gameplay emerge in the market, as long as they involve private key management and TEE verifiable consensus, Phala's TEE environment is the best solution. After the upgrade of the token model, $PHA will also become the golden shovel for the Agent + TEE sector.

Phala is About to Upgrade Its Token Economic Model to Create a Token Flywheel for More TEE Use Cases

Phala has experienced multiple bull and bear markets and currently maintains a business model focused on Intel SGX in its token economic model. According to Paradigm's article "The 5 Levels of Secure Hardware," there are five levels of secure hardware, with the second level referring to: slightly lower performance but better developer experience, allowing for more expressive applications, with no improvement in security. Intel SGX falls into this level, specifically serving TEE apps. As mentioned at the beginning of this article, sensitive locally stored data such as fingerprint enrollment and comparison, as well as facial recognition on computers and smartphones, utilize Intel SGX. This previous generation of TEE is specifically designed for app services.

Source: Paradigm

As use cases further expand, they are no longer limited to the application level but rise to the system level. Intel SGX cannot meet market demands, leading to the emergence of Intel TDX. Intel TDX is designed specifically for virtual machines, and even NVIDIA's H100 and H200 have begun to support TEE, which serves AI.

Source: Paradigm

Returning to Phala, although it has already taken the lead in supporting the third level, the $PHA token economic model and mainnet are still designed around Intel SGX from 4 to 5 years ago. Therefore, even though Phala has collaborated with numerous Web3 protocols in terms of products and practical use cases, the token model has not been updated in sync, and the corresponding flywheel cannot yet operate. Thus, the current state of revenue and product is not aligned. However, this state will not last long, as Phala will soon upgrade its token model and mainnet to match the stage of Intel TDX and NVIDIA GPUs.

Additionally, Phala will enhance the value capture capability of $PHA. In the future, the latest Agents launched on Spore.fun will airdrop tokens to $PHA holders, officially transforming it into a golden shovel.

TEE itself is not a new technology, but with the emergence of AI Agents as a new landing scenario, market discussions have begun to rise. Phala is not a so-called "fast pass" brought about by the emotional explosion on PumpFun; its value growth is based on years of deep product accumulation, leading to a significant breakthrough. Agent + TEE is not a fleeting trend that comes fiercely and leaves nothing behind; rather, it is fertile ground that allows more Agent landing scenarios to take root and thrive.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。