The combination of AI and blockchain and related legal risks

IntermediateMar 27, 2024
The core of the AI technology revolution lies in ample computing power, algorithm models, and a vast amount of training data. Currently, high-performance GPU computing power is in short supply and expensive, algorithms tend to be homogenized, and there are issues regarding data compliance and privacy protection for model training data. The decentralized and distributed storage characteristics of blockchain technology can facilitate its integration with AI.
The combination of AI and blockchain and related legal risks

In recent years, with the successive release of GPT series products, artificial intelligence is transforming various industries. We have witnessed various AI applications entering our daily work and life, enhancing work efficiency, changing lifestyle habits, and reducing enterprise operating costs. We must admit that AI is becoming the starting point for the next technological revolution.

The core of the AI technology revolution lies in ample computing power, algorithm models, and a vast amount of training data. Currently, high-performance GPU computing power is in short supply and expensive, algorithms tend to be homogenized, and there are issues regarding data compliance and privacy protection for model training data. Blockchain technology possesses characteristics such as decentralization and distributed storage, which can be effectively applied in the development, deployment, and operation of AI models.

Utilizing The Characteristics Of Blockchain To Solve The AI Computing Power Issue.

For the AI development process, facing issues such as GPU computing power shortages and high usage costs, some blockchain projects are attempting to address them through blockchain-based solutions.

Render Network is a high-performance distributed rendering platform that bridges the gap between GPU computing power demand and idle GPU resources providers using industry-leading otoy software. This setup allows idle GPU resources to be supplied to high-demand computing fields such as artificial intelligence and virtual reality at a lower cost.

In this ecosystem, providers of idle GPUs connect their devices to Render Network to complete various rendering tasks, while demanders compensate GPU providers with token rewards. This decentralized approach maximizes resource utilization efficiency, creates value for participants, and reduces the development and operational costs of artificial intelligence. In December last year, Render achieved a significant technological leap by migrating its infrastructure from the Ethereum chain to the high TPS Solana, leveraging Solana’s high performance and greater scalability to enhance Render’s capabilities, including real-time streaming and state compression.

Rendered Image on Render Network

Akash is a decentralized computing platform that aggregates idle CPU, GPU, storage, bandwidth, dedicated IP addresses, and other network resources worldwide and rents them out to enterprises and individuals requiring high computational power for tasks such as artificial intelligence. This allows them to fully leverage these resources and provide GPU rental services. Users who provide GPU rental resources receive AKT tokens, while demanders gain access to computing power at low costs. The platform token AKT is not only used for settling payments for rented network resources but also serves as an incentive to encourage validators to participate in ecosystem governance and network security maintenance. The platform charges a certain transaction fee for settling payments for network resources, allowing all participants in the platform ecosystem to generate income and driving the platform’s long-term viability and continuous growth of its business model.

Akash Network’s real-time statistical graph of network resources

Livepeer is a video infrastructure network platform for live and on-demand streaming media. Users can join the network by running the platform software and utilize their computer’s GPU, bandwidth, and other resources for transcoding and distributing videos. This approach improves the reliability of video streams and reduces related costs such as transcoding and distribution by up to 50 times. Furthermore, the Livepeer project is introducing AI video computing tasks to the Livepeer network, utilizing its GPU network operated by orchestrators to generate high-quality AI videos, thereby reducing the cost of creating video content.

From the description of blockchain projects above, it’s evident that blockchain can leverage its decentralized and distributed characteristics to effectively utilize idle network resources to address the current shortage of AI computing power and high costs. Once this model is validated and recognized in more real-world scenarios and by AI startups in the future, it will significantly alleviate the computing power issue.

The Integration of AI with Blockchain Data.

Data is the foundation of AI models, and the data used to train models determines the differences between various AI models. Compared to other data sources, blockchain data is of higher quality and transparent, enabling the identification of users on the blockchain.

Arkham is a platform that rewards users for providing on-chain data and intelligence analysis using AI technology. Its proprietary artificial intelligence engine, ULTRA, can label on-chain addresses with real-world users, allowing decentralized on-chain anonymous addresses to be linked to real-world individuals. By obtaining a large amount of labeled data for on-chain anonymous addresses through AI models, users can uncover on-chain transaction information of entities through Arkham. It is well known that the biggest challenge in investigating cryptocurrency crimes is identifying fund transfers through anonymous addresses. Regulatory authorities can trace and investigate criminal activities such as money laundering and fraud through the label data provided by Arkham.

On-chain data visualization map of Arkham platform

In addition, Arkham also features an on-chain intelligence information trading function. The inter-exchange feature of Arkham enables the exchangeability between on-chain addresses and off-chain real-world information. Users can gather on-chain information intelligence through bounty rewards on the platform, and valuable on-chain information can also be auctioned on the platform. For a detailed analysis of the specific products, you can refer to the article “Can Arkham Become a Powerful Tool for On-Chain Regulation?“ previously written.

Arkham’s artificial intelligence engine, ULTRA, received support during development from Palantir, a big data analytics and intelligence services company that provides artificial intelligence services to the U.S. government, as well as from the founders of OpenAI. With such strong support and access to a powerful AI model training data source, Arkham possesses the most robust on-chain data labeling library in the industry.

Addressing the high cost of storing large amounts of data for AI model training, blockchain storage projects such as Arweave, Filecoin, and Storj have provided solutions. Whether it’s Arweave’s one-time payment for permanent storage or Filecoin’s efficient pay-as-you-go model, these solutions significantly reduce data storage costs. Additionally, decentralized storage can mitigate the risk of data loss due to natural disasters compared to traditional storage methods.

While using ChatGPT can improve work efficiency, optimizing the model to enhance the accuracy of AI conversations requires large amounts of user data for training and fine-tuning. Therefore, there is a risk of sensitive data and personal privacy data leakage. Zama is an open-source cryptography company that builds state-of-the-art fully homomorphic encryption (FHE) solutions for blockchain and artificial intelligence. Zama Concrete ML can securely handle sensitive data, enable data collaboration between different institutions while maintaining confidentiality, improve efficiency and data security. It encrypts privacy data such as personal medical records during training, ensuring that each user can only see the final result and not other people’s sensitive data.

The Integration of AI agents with Blockchain Projects.

OpenAI defines an AI Agent as a system driven by a large language model (LLM) that possesses the ability to autonomously understand, perceive, plan, remember, and use tools, enabling it to automate the execution of complex tasks. With the successive releases of OpenAI’s GPT models, there are increasingly more applications of AI agents being implemented. is a self-learning blockchain network primarily facilitating economic activities among autonomous AI agents. consists of four parts: AI Agents, Agentverse, AI Engine, and Fetch network. Users can create, develop, and deploy their own AI agents using the AI agent use cases provided by the platform on Agentverse. They can also promote their AI agents to other users on the platform. DeltaV is the AI-based chat interface in Users input requests through this interface, and the AI Engine reads the user input, converts it into actionable tasks, and selects the most suitable AI agent in Agentverse to execute the task. Currently, the German company Bosch is collaborating with to research the integration of AI agent technology with mobility and smart home applications, jointly opening the door to the Web3 era of the Internet of Things economy.

The composition of the ecosystem

In addition, the AI ​​Agent application QnA3.AI introduces the encryption industry’s AI question and answer robots, technical analysis robots and asset trading capabilities into the Web3 world. Through QnA3 Bot, users can collect, analyze and execute actual transactions when trading crypto assets. Behavior is realized through the product functions of “Question and Answer”, “Technical Analysis” and “Real-time Trading”, which minimizes the interference of users’ subjective emotions in their trading decisions.

Possible legal risks

1. Data export risks

In the above introduction, it is mentioned that some decentralized storage projects are addressing the data storage issue for AI model training at a lower cost. This lowers the barrier for individuals and startups dedicated to AI entrepreneurship. However, this decentralized storage approach may pose risks of data leaving the jurisdiction.

The National Internet Information Office has issued the “Guidelines for Security Assessment of Data Export (First Edition)”, which clearly states that data export behavior includes:

(1) Transferring and storing data collected and generated during domestic operations to overseas locations by data processors;

(2) Storing data collected and generated by data processors domestically, and allowing institutions, organizations, or individuals overseas to query, retrieve, download, and export the data;

(3) Other data export behaviors as regulated by the National Internet Information Office.

So, what is the definition of “export”? Article 89 of the Exit and Entry Administration Law of the People’s Republic of China clearly states that “export” refers to traveling from mainland China to other countries or regions, traveling from mainland China to the Hong Kong Special Administrative Region or the Macao Special Administrative Region, and traveling from mainland China to Taiwan. Therefore, it can be seen that the determination of whether there is an export is based on jurisdiction.

For decentralized storage projects, users store data in decentralized distributed networks such as IPFS. The files stored in the network are divided into several small chunks of data, encrypted, and stored in various nodes, with storage nodes distributed worldwide. Imagine if a domestic AI startup were to store the data for training AI models on nodes of such decentralized projects, there would indeed be a risk of data export.

2. Risk of sensitive privacy data leakage

In AI Agent applications like QnA3.AI, users engage in conversations with AI to obtain trading information for encrypted assets and execute transactions. The personal Q&A dialogue generated from these interactions poses a risk of privacy data exposure if utilized by the project for model training and optimization. Such leakage of transaction data, if exploited by malicious actors, could lead to investment failures and potentially greater losses.


  1. This article is reprinted from [web3caff], All copyrights belong to the original author [Chris Chuyan]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
  2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
  3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.
Start Now
Sign up and get a
Create Account