Article by imToken
In recent times, Vitalik has mentioned several times a term that seems somewhat unfamiliar: CROPS.
The systematic emergence of this concept can be traced back to March 13, when the Ethereum Foundation Board released the "EF Mandate" document, explicitly stating its primary focus on Ethereum’s censorship resistance, openness, privacy, and security—collectively known as CROPS—to serve user self-sovereignty while maintaining extractability resistance and a more seamless user experience.
This statement is actually very important, especially as AI begins to enter wallet and automated execution scenarios—CROPS will no longer be limited to issues of Ethereum’s values, but could instead become a question of whether users can continue to control their own digital lives in the AI era.
What exactly is CROPS?
To understand CROPS, first move beyond a common misconception: while Ethereum certainly needs to improve performance and reduce costs, it’s not merely about competing with other blockchains to see who is faster or has lower fees.
Although speed and cost may seem the most immediate benefits from a short-term user experience perspective, over the past two years, Ethereum’s stance has become increasingly clear: it aims to provide a more foundational set of capabilities—allowing users to hold assets, express identity, sign transactions, and participate in coordination without relying on a single platform, surrendering ultimate control, or being arbitrarily blocked by a centralized service.
This is what CROPS means.
Within the context of the EF Mandate, CROPS primarily refers to five key areas, which are also the acronyms for: Censorship Resistance, Capture Resistance (this was later added by Vitalik), Open Source, Privacy, and Security—namely, resistance to censorship, resistance to capture, open source, privacy, and security.
- C - Censorship Resistance: Ensure that transactions and smart contracts are immutable and cannot be terminated due to pressure from any external political or centralized entity;
- R - Capture Resistance: Prevents Ethereum's governance, development roadmap, and critical access points from being long-term controlled by a small group of interested parties;
- O - Open Source / Openness: Maintain fully open-source code and ensure absolute freedom of access to the ecosystem;
- P - Privacy: Using cryptographic techniques to protect users' right to confidentiality on a transparent ledger;
- S - Security: Uphold the foundational standards to deliver unbreakable, ultimate settlement security;
Taken together, these items form a clear set of selection and guidance principles that align closely with Ethereum’s longstanding values.
For example, at the protocol level, Ethereum must continuously improve censorship resistance, client diversity, validator decentralization, and formal verification; at the application level, wallets, RPC services, browsers, signing interfaces, and account systems must reduce reliance on centralized entry points; at the user experience level, security cannot depend solely on users understanding complex transactions—they must be supported by clearer signing displays, more verifiable interactions, and more comprehensive risk warnings that surface risks before actions are taken.
This is also why EF has recently advanced several more specific initiatives around security, privacy, protocol resilience, and ecosystem public goods—such as the Ethereum Audit Subsidy program, which aims to lower the barrier for Ethereum ecosystem developers to access high-quality security audits. More broadly, this is not merely about subsidizing costs, but about making “security” accessible to more small and medium-sized developers, moving it beyond a high-cost service only affordable to a few large projects.
In late May, Vitalik once again shared his views on the future direction of the EF, emphasizing that the EF should become a smaller, more clearly positioned, and more focused organization centered on long-term sustainability, rather than attempting to address every need within the ecosystem. The reason is pragmatic: the EF does not have unlimited resources nor a consistent revenue stream from staking or transaction fees, and therefore should prioritize its limited resources toward tasks that are critical to Ethereum’s realization of CROPS values and that other entities cannot reliably undertake.
In other words, at this pivotal historical stage of Ethereum’s transition, CROPS is not an abstract slogan prioritizing ideals over reality, but rather an external framework that defines and constrains what the EF should and should not do.
II. When CROPS Meets AI: The Convergence of Two Parallel Universes
Vitalik Buterin most recently brought CROPS into broader discussion within the context of AI.
On May 28, Vitalik Buterin posted an update on his localized AI progress, stating that a 2-bit quantized version of DeepSeek V4 has been released, capable of running within approximately 90 GB of VRAM, achieving speeds of about 35 tok/s on Apple hardware and around 7 tok/s on AMD hardware. He added that true "CROPS AI" should support multiple hardware platforms, not just "decentralized AI."
He also noted significant overlap between the CROPS Ethereum access layer and CROPS AI, such as paid remote LLM calls via zero-knowledge proofs and private RPC reads on Ethereum. In the future, more AI models fine-tuned for Ethereum-specific use cases should emerge to enhance the security of smart contracts, protocol code, and the broader ecosystem.
This actually places Ethereum and AI within the same problem framework.
In the past, when we discussed AI, we often focused on model capabilities—such as whether it could write code, especially whether it could replace humans in handling complex tasks. But from a user security perspective, the real change brought by AI is not just "greater capability," but that it is transforming the entry points for digital operations.
As before, applications used to have clearly defined interfaces: we opened a wallet to transfer funds, opened a DApp to trade, opened a browser to search, and opened a social platform to post—each app had distinct boundaries. But with the emergence of AI agents, these boundaries are becoming increasingly blurred; users no longer click through individual functions, but instead express their intentions in natural language:
Find the optimal cross-chain path for me, execute a swap for me, organize my assets, activate a specific DeFi strategy, and generate and send a transaction...
This sounds convenient, but it also raises a more important question: when AI becomes your digital agent, what transactions is it signing on your behalf, and what personal information is it exposing?
If AI operates entirely on centralized cloud servers, users' asset information, transaction intentions, address relationships, identity preferences, and behavioral patterns could become concentrated in the hands of a few service providers—especially when executing on-chain operations that rely on opaque APIs, centralized RPCs, black-box plugins, and unverifiable reasoning processes. While users may gain greater convenience, they may also find it harder to know exactly what they have surrendered.
This is the question CROPS AI aims to answer.
An AI better aligned with CROPS should not only be powerful but also as censorship-resistant, open, privacy-preserving, and secure as possible. It should ideally run locally and, in sensitive scenarios, minimize reliance on centralized cloud services, reduce information leakage, and ensure users understand, confirm, and retain ultimate control.
In other words, AI cannot merely be a smarter black box—especially in Web3 contexts, where AI may soon do more than summarize articles, write code, or handle customer service; it could directly participate in asset management and automated execution.
The closer it is to user assets, the more important CROPS becomes.
This is also why the CROPS Ethereum access layer and CROPS AI intersect.
III. What Web3 incremental opportunities can be explored in this intersection?
From this perspective, it’s very natural that Vitalik recently mentioned an intersection between the CROPS Ethereum Access Layer and CROPS AI.
Because whether it’s Ethereum or AI, the core issue users face is becoming the same—how can I use AI assistance without fully surrendering my privacy, identity, assets, and autonomy to centralized intermediaries?
- On Ethereum, this manifests as: How do users access on-chain data? How do they connect to RPC? How do they sign transactions? How do they verify whether DApp interactions are secure? How can they avoid having all wallet queries, balance checks, and transaction broadcasts go through a few centralized services?
- On the AI side, this manifests as: How do users invoke models? How can we ensure prompts and personal data are not misused? How can local models handle sensitive tasks? And how can we minimize exposure of our identity and intent when remote large models are required?
These two sets of questions may seem different on the surface, but they are fundamentally similar.
For example, when Ethereum users check their balances, read transaction histories, or simulate transaction outcomes, they often rely on RPC services. Although RPC appears to be merely a technical interface, it can collect information such as your IP address, wallet address, query patterns, asset structure, and interaction pathways. If this data is centrally aggregated, users’ on-chain privacy can be gradually reconstructed.
However, when AI users invoke remote models, they may also expose their preferences, financial information, or even identity clues; if users later use AI to handle wallet operations, these risks will be further amplified.
So what Vitalik mentioned—ZK-paid remote LLM calls and private Ethereum RPC reads—essentially aim to solve the same problem: how to access remote services without exposing all of your information.
This is where CROPS Ethereum and CROPS AI intersect: on one side, a blockchain-based access layer that is more private, verifiable, and requires fewer trust assumptions; on the other, an AI execution environment that is more open, localized, and secure. Together, they may form a new gateway for users to enter the digital world.
Extending from the underlying logic of CROPS, the entire Web3 ecosystem—particularly the wallet layer as the traffic entry point—will undoubtedly take on more roles:
When users begin expressing on-chain requests in natural language, a wallet is no longer just a signing tool—it becomes the control center for the user’s digital actions, requiring it to help users determine whether this DApp can be connected, what exactly will happen with this transaction, and whether this AI agent is calling for unnecessary data.
From this perspective, CROPS is not an abstract value but directly influences the design direction of wallet products and drives the next decade’s shift toward integrating Web3 interaction experiences within the wallet ecosystem.
In conclusion
Although, under current market conditions, many people may have less interest in purely conceptual projects.
But the colder the market, the more likely it is that people will overlook technical factors which may not seem exciting in the short term but ultimately determine the long-term direction.
CROPS is worth paying attention to not because it created a new trend, but because it reframes the long-standing issues of Ethereum and AI within a single framework: as digital systems grow more powerful, can users still maintain control over them?
After all, security and privacy cannot be mere afterthoughts.
From this perspective, in an era where AI is rapidly taking over the digital world, the true positive factor for Ethereum’s continued relevance and use may lie in its ability to adapt and thrive.
In an era where AI is rapidly taking over the digital world, Ethereum’s true value lies in being more understandable, verifiable, private, and secure.