Decentralized Artificial Intelligence (DeAI) is becoming an increasingly important part of AI infrastructure as the demand for safer and more efficient AI computing grows. However, a new whitepaper suggests that DeAI needs the integration of Trusted Execution Environment (TEE) technology to fully unleash its potential.
“TEE’s role will become increasingly important as the demand for safer and more efficient AI computing grows,” said Nukri Basharuli, the co-author of the whitepaper and CEO of Super Protocol, in an interview with Decrypt. “Their seamless integration with existing blockchain architectures makes them indispensable in developing future-proof DeAI systems.”
The current landscape of AI relies on reliable private user data to deliver high-quality results. However, sharing data and models in the AI community poses risks such as confidentiality and ownership infringement.
DeAI promises to create a collaborative environment that limits access to sensitive data for all participants. However, the authors of the whitepaper argue that the current state of this new technology has limitations in terms of scalability and reliability, which hinder its potential.
The biggest advantage of blockchain lies in its ability to transparently validate transactions and smart contracts. However, David Attermann, Web3 and AI Investment Director at M31 Capital, a global investment firm specializing in crypto assets and blockchain technology, explains that traditional methods are insufficient when it comes to complex computations required by AI. He states, “Using blockchain alone to verify these computations is extremely expensive and impractical.” This limitation severely hampers the scalability and efficiency of DeAI.
So, what is TEE? TEE is embedded in CPUs and GPUs, providing a secure and isolated environment for executing code and processing data. This isolation ensures that data and computations within the TEE remain protected even if the main system is compromised.
In fact, the device you are reading this on most likely includes a TEE. They are used to store sensitive information like biometric data in smartphones and computers, as well as cryptographic keys in hardware wallets for cryptocurrencies.
“TEE is the cornerstone of many security applications today,” says Basharuli. He points out that their integration with blockchain systems “can transform them from merely ‘trusted’ to truly ‘untrusted’ by leveraging cryptographic primitives and consensus mechanisms.”
He further explains how platforms like Super Protocol demonstrate the effective integration of TEE with blockchain technology in DeAI applications. By combining smart contract governance with clustered TEE computing capabilities, they achieve trustless and secure AI computing.
“In an untrusted environment, collaboration becomes untrusted,” Basharuli states. “Everyone can trust that this system can handle their assets safely and confidentially.”
The authors of the whitepaper explain that TEE brings several key benefits to decentralized AI applications. Attermann emphasizes that TEE provides a crucial level of security for sensitive AI computations, ensuring data remains encrypted and protected throughout the entire process.
Basharuli adds that by isolating sensitive data from the wider system, TEE ensures data remains confidential and secure even in the face of threats to the broader system.
TEE also maintains the integrity of data and code, preventing unauthorized modifications. Attermann states, “Integrity is crucial, and TEE maintains it by isolating the execution environment.”
Additionally, TEE provides an isolated execution environment that ensures secure processing of sensitive operations and enables cryptographic proofs to verify that the code running in the secure enclave is genuine and unaltered.
TEE is positioned alongside other secure computing technologies such as Zero-Knowledge Proofs (ZKPs) and Fully Homomorphic Encryption (FHE). However, Attermann explains that these technologies have limitations that make them unsuitable for DeAI in their current state, including high costs and low performance.
“At the intersection of TEE and blockchain, we ensure security and confidentiality while enhancing scalability and performance,” Attermann tells Decrypt. “This makes TEE the most practical and direct solution for achieving verifiable, untrusted AI computing.”
“TEE is very cheap and very fast,” he adds, “but for some reason, they are not sexy. We are working to change that.”
Integrating TEE into DeAI systems still faces some challenges. The whitepaper argues that ensuring the correct and consistent use of TEE in decentralized networks requires strong governance and oversight. It is also crucial to foster industry-wide acceptance and understanding of TEE technology.
Basharuli states that Super Protocol and other pioneers in the field are actively addressing these challenges. By demonstrating the practical benefits of TEE in real-world applications and collaborating with industry leaders like Nvidia, they are paving the way for wider adoption. He adds, “Confidentiality is crucial for secure collaboration in AI, and cryptocurrency incentivizes collaboration, accelerating adoption.”
The whitepaper concludes that TEE provides the necessary security, confidentiality, and verifiability for DeAI to fully realize its potential. “By integrating TEE with blockchain, we can go beyond ‘trusted’ systems and enter a truly ‘untrusted’ environment, ensuring a decentralized, secure, and scalable future for AI.”
For developers eager to start experimenting with DeAI, the Super Protocol AI Marketplace is now live. It allows collaboration on AI models and applications in a confidential and non-disclosing computing environment, leveraging everyone’s solutions, storage, data, or computing power.
Soon, developers will also be able to monetize their AI agents and applications on the Super AI Marketplace using transparent blockchain tools that distribute income among all contributors. They will also use stablecoins for payments as part of broader efforts to make it easier for Web2 developers to integrate into the ecosystem.
Basharuli tells Decrypt, “Building a confidential AI collaboration hub requires solving the chicken-and-egg problem of supply and demand, which means users should bring useful datasets and models before starting to develop AI products.” Super Protocol’s solution is to provide a comprehensive incentive program to attract models, data, and users into the AI marketplace in exchange for token-based rewards.
For more information about Super Protocol, please visit their website, Twitter, and LinkedIn. For more information about M31 Capital, please visit their website, Twitter, LinkedIn, and Substack.
