Original | Odaily Planet Daily (@OdailyChina)
Author | DingDang (@XiaMiPP)
The hot topic OpenClaw has started endorsing Privacy AI, and the "desperate crypto retail investors" seem to have found a new direction for speculation.
It is precisely within this narrative context that a batch of projects related to privacy computing and AI Agent infrastructure have begun to re-enter the market's view. Odaily Planet Daily's review found that during this wave of heated discussion, several projects have already become potential beneficiaries.
VVV(#133)
Venice is an AI generation platform focused on censorship resistance + privacy, positioning itself as a decentralized version of ChatGPT. The starting point of the Privacy AI hype originated from Venice. This is because OpenClaw had once highlighted Venice in its official documentation, only to remove it hastily within 24 hours. Although the recommendation could be removed, this action drew more attention to Venice and its privacy-first features.
Unlike most AI projects, Venice's core narrative is not AI model capability, but privacy itself. Against the backdrop of mainstream AI platforms increasingly strengthening content censorship, and with ongoing controversies over AI data leaks and model training, this "no recording, no censorship" product positioning precisely hits the most sensitive values of the crypto community.
In an era where the AI Agent trend is rapidly fermenting, Venice has stepped into this "era dividend" quite aptly. More coincidentally, the Venice project is actively reducing the token supply of VVV, decreasing inflation. Increased demand meeting reduced supply further strengthens the positive feedback expectation for the VVV token.
Read reference: OpenClaw Endorses Venice.ai, VVV Token Surges Over 500% in a Month
NEAR(#43)
Near Protocol, this veteran high-performance public chain project, is also actively seeking self-reinvention under the impact of the AI wave. It is no longer just a "traditional L1" pursuing TPS and low gas, but is gradually shifting its narrative focus towards the execution layer and settlement infrastructure in the era of AI Agents, attempting to find a new growth narrative in this new technological cycle.
Since the beginning of 2025, it has been vigorously promoting the NEAR Intents (Intent System). This system allows users or AI agents to simply express the "final desired outcome," and the backend will automatically complete complex operations across 35+ chains without manual bridging, wallet switching, or routing management.
On February 25, 2026, NEAR officially upgraded this intent system and launched Confidential Intents. This version introduces privacy computing capabilities into the original intent execution framework. Through Near's privacy sharding mechanism combined with Trusted Execution Environments (TEE), cross-chain transactions can hide key details during execution, such as swap paths, trade size, or specific strategies. However, it does not enforce privacy on all transactions like Zcash or Monero, but rather adds an optional layer of privacy protection to intent execution. Its main goal is not to anonymize transactions, but to prevent MEV, front-running, and sandwich attacks and other on-chain arbitrage behaviors, thereby making transactions more secure during execution.
In the future, AI agents might become the main "users" of blockchain. They will autonomously hold assets, conduct cross-chain transactions, execute strategies, and even coordinate with each other. Under this vision, blockchain not only needs to handle high-frequency transactions but must also provide capabilities like verifiable execution, privacy computing, and cross-chain coordination.
Near's current layout is precisely built around this imagination. It attempts to build an open network that can both support AI agents in automatically executing complex tasks and ensure the process is verifiable and secure. Against the backdrop of the continuously impacting AI wave, this transformation can be seen both as an attempt to actively embrace the new narrative and as a self-reshaping of a veteran public chain in the new cycle.
SAHARA(#295)
Sahara AI's core goal is to build a decentralized, transparent, and secure AI ecosystem, making the development, training, deployment, and commercialization of AI fairer and more trustworthy. The project is committed to solving the problems currently faced by the AI industry, such as data privacy, algorithmic bias, and unclear model ownership.
The rise of AI Agents is bringing forth a new question: Who exactly do the data, models, and capabilities used by these Agents belong to? In the current AI industry structure, this question hasn't been properly resolved. The data required to train models often comes from a large number of dispersed contributors, but the final profits are highly concentrated in the hands of a few AI companies; model developers, even with technical capabilities, often can only依附 on platform ecosystems; and as AI Agents begin to call models, data, and tools autonomously, the entire value chain will become more complex. Without a clear set of rights confirmation and profit-sharing mechanisms, the future AI economy will likely repeat the path of Web2: data owned by users, value captured by platforms.
Sahara AI is precisely trying to establish new rules in this环节. Its ClawGuard security system provides verifiable safety guardrails for AI agents, ensuring they operate safely within preset rules. The Data Service Platform (DSP) allows users to earn token incentives by labeling and contributing AI training data, gradually forming a decentralized data market. Under this mechanism, data contributors can not only participate in the AI model training process but also receive continuous收益 when the data is used, while the platform also ensures data quality and privacy protection through on-chain mechanisms.
PHA(#601)
Phala Network is a privacy-preserving smart contract platform built on Substrate, aiming to provide verifiable privacy-preserving computation services for Web3 applications. To understand why Phala benefits from the AI Agent热潮, one first needs to answer a more fundamental question: What infrastructure does the operation of AI Agents actually rely on?
If we break down the current Agent ecosystem, its tech stack can be roughly divided into several layers. The top layer is the model layer, which includes various large language models or reasoning models, such as OpenAI, Claude, and a series of open-source models; below that is the Agent framework layer, including tools like LangChain, AutoGPT, OpenClaw, etc., which are responsible for organizing tasks, scheduling models, and calling external tools; further down is the execution environment layer, where the Agent actually runs code, calls APIs, and executes automated tasks; additionally, there is the payment and identity layer, used to handle payments, identity, and reputation systems between Agents; and at the very bottom is the compute power and privacy layer, responsible for ensuring the computation process is trustworthy and data security is not compromised.
From this structure, Phala's position恰恰 spans the execution environment layer and the compute power & privacy layer. Its core technology—a confidential computing network based on TEE (Trusted Execution Environment)—enables AI Agents to run programs securely off-chain while ensuring the computation process is verifiable and data is not窥探 by external parties. This is particularly crucial in the Agent economy.
In terms of specific ecosystem implementation, Phala has already begun to integrate with AI Agent projects. For example, Phala collaborated with ai16z to build a TEE component for its Eliza multi-agent framework, integrating trusted execution technology directly into the Agent runtime environment; meanwhile, some AI Agent token issuance projects (like aiPool) have also adopted Phala's TEE technology to manage private keys and on-chain assets.
In the future, as AI Agents evolve from "chat tools" into digital entities capable of holding funds, executing transactions, and even operating protocols, secure execution environments will gradually become an indispensable infrastructure layer for the entire Agent ecosystem, and Phala is trying to occupy this position.
Conclusion
When reviewing these projects, an interesting finding is: The time when these tokens actually started rising was actually earlier than the recommendation event these past few days. That is to say, before Venice brought "Privacy AI" to the forefront, a portion of funds in the market had already noticed this direction in advance, it just lacked a clear enough narrative trigger point at the time. The OpenClaw recommendation event was just a fuse that ignited attention.
In fact, whether it's a16z or Delphi Digital, their annual investment research reports for 2025 listed privacy and AI as key focus areas for 2026. However, when these macro judgments真正落到 the market, they often need a specific event to trigger consensus. And in early 2026, privacy and AI have come before us in such a combined form.
As for whether this will become the next long-term trend or just another short-lived thematic speculation,恐怕 still requires time to provide the answer.
