jshelby

Author:  Nukri Basharuli, Founder and CEO, Super Protocol

True Web3 Games, with their potential for rich gaming experiences, advanced AI agents, and genuine digital asset ownership, can only reach their full potential through the implementation of Confidential Computing in a truly decentralized manner. The Confidential Computing Consortium, alongside its member Super Protocol, is at the forefront of this revolution, demonstrating how these technologies can unlock new business opportunities.

Super Protocol serves a dual role in the evolving digital landscape. As a confidential and self-sovereign AI Cloud, it focuses on decentralization, privacy, and verifiability. Its computing network of commonly adopted types of GPU and CPU operates in a confidential mode under the orchestration of Smart Contracts on the Polygon blockchain. This makes Super a decentralized alternative to centralized clouds like Amazon AWS for Web3 AI projects. Additionally, as an AI Marketplace, Super Protocol differs from traditional AI marketplaces like Hugging Face by offering AI models and Data owners the unique ability to share and monetize their assets in a fully confidential, self-sovereign mode. The value of Super Protocol is well illustrated by the examples of its clients.

In this blog, we’ll explore how Super Protocol’s AI cloud is set to transform the gaming industry, enabling a secure and self-sovereign experience that could redefine the future of digital entertainment.

Example: Veriplay 

About Veriplay 

Veriplay is a startup that is developing a gaming platform compatible with Immutable and Polygon. This platform will enable creating AI agents in Web3 Games that can be traded on the open market as dynamic NFTs. 

The Veriplay team, with a proven track record of working with industry giants like Playrix, Warner Brothers, Google, and Crytek, is on a mission to revolutionize the gaming landscape by introducing verifiable and tradable AI agents. 

This innovative approach aims to address the limitations of traditional gaming experiences and empower players with unprecedented control over their in-game assets. 

Super Protocol and Veriplay research and testing efforts have utilized NVIDIA’s H100 GPUs, provided by the NVIDIA Confidential Computing team (more details in Super Protocol Press-release: 

https://www.linkedin.com/posts/superprotocol_confidentialcomputing-depin-nvidiainception -activity-7169336537371914242-LBV3?utm_source=share&utm_medium=member_desktop 

Project Goal 

Veriplay’s goal is to give players the ability to truly own AI game agents. To do so, it aims to develop a reliable, Web3-compatible gaming platform for integrating verifiable tokenized AI agents into games with the following characteristics: 

• Player AI Models are Protected from Unauthorized Alterations: Veriplay wants to protect player AI models from any unauthorized modifications, whether initiated by the game developer or external malicious actors. 
• AI Model Training is Verifiable: This means that it is possible to verify how the AI agents were trained, which guarantees their fairness and transparency. 
• Decentralization (Smart Contract Orchestration): Smart contracts will govern the execution

of AI computations and data storage, ensuring transparency and immutability, and eliminating human administration layer. 

• Free trading of AI NFTs on marketplaces : Veriplay revolutionizes AI agent ownership by transforming them into tradable digital assets managed by players through dynamic NFTs. 

It is evident that without AI computation privacy, verifiability of AI model training history, and smart contract management, the integrity of AI agents as digital assets will be irrevocably compromised, shattering market trust and leading to market rejection of such assets. 

The Centralized Infrastructure Problem 

There are several problems with creating trusted AI agents in centralized infrastructure, such as Amazon Web Services (AWS) or Google Cloud: 

• Difficulty of Verification: It is difficult to verify that AI agents have been trained and operate according to the rules of the game declared by the developer. This is especially important when AI agents become tradable assets or when they are used in competitions with prize money. 
• Risk of Developer Manipulation: Developers have the ability to alter or duplicate an AI agent trained by a player who has invested time and money into the training process. For instance, a developer could duplicate a successful model that frequently wins competitions and sell it to other players as this developer’s original creation. 
• Player’s Inability to Own Agents: In centralized AI agent infrastructures, players lack true ownership of their agents, being confined to developer-defined capabilities and pricing models. While creating a simple NFT for AI agent ownership partially addresses this, it falls short of true self-custodial ownership. For this type of ownership to be achieved, the AI NFT must be dynamic, linked to all AI agent components, maintain security and verifiability, and prevent human administration access – all impossible within centralized frameworks. 

To sum it up, centralized infrastructure poses significant risks that not only diminish the value of players’ time and investments in training their agents, but also severely restrict monetization opportunities for both players and game studios. 

Conversely, with the implementation of trustless AI Agents in games, both players and developers could generate additional income by trading dynamic AI NFTs on marketplaces, renting out Agents to each other, ghosting and participating in the championships with cash prize funds, and more. 

Super Protocol and Veriplay Solution 

In contrast to AI agents confined within centralized infrastructures and under the complete control of game developers, Web3-compatible gaming agents powered by Super Protocol and the Veriplay platform will exhibit the following advantages:

1. Confidentiality and Sovereignty of AI Agent – players retain exclusive sovereignty over their AI agents, encompassing models, data, and computational resources, effectively eliminating the possibility of third-party manipulation. 

• Confidential Enclave Technology: Web3-compliant AI agents are computed in confidential enclaves. Confidential enclaves operate based on the Trusted Execution Environment (TEE) technology supported by Nvidia H100 GPU chips. TEE allows creating a secure area inside the processor for safe storage, processing, and protection of confidential data. Even physical access to the server will not grant access to the applications running in the enclaves. No one except the owner of the Agent knows on which servers their data is being processed, as TEE ensures complete isolation of sensitive data and models. 
• Access and control over the system are only granted to the smart contract and verified applications loaded into it. The computational resources used for the Agent’s operation are automatically authenticated by the protocol, ensuring the user that they are processing their model and data securely. By design, these resources cannot be tampered with or exploited maliciously. The owner alone manages the model, data, and interactions of their gaming agent. 

As a result, users can be confident that the game developer cannot alter or copy the model since they do not have access to it. 

2. Verifiability of AI Agent training and game interactions – maintaining the verifiability of the Agent throughout the chain from the storage to the server is guaranteed by the following functions: 

• The client application and the server are mutually authenticated on a TLS connection. They exchange messages signed with a secret key. Messages contain information about the hardware, application, and its settings. 
• After mutual authentication, the game application computing process initiates. The outcome will be signed using the enclave key, maintaining the chain of trust. This trust continuity prevents unauthorized alterations, ensuring players can rely on the computation’s result. 

Therefore, the verifiability of the AI Agent’s track record and its immutable nature reassure players that acquiring an AI Agent guarantees possession of an asset with the promised properties. Moreover, by investing in its continuous training, they can have confidence that the future market price of the Agent will accurately reflect their training endeavors. 

3. The decentralization and removal of human administration are achieved through orchestration by a smart contract system. Via smart contracts, Super Protocol entirely separates the gaming process from server and cloud owners, guaranteeing trust, flexibility and reliability. 

• Smart Contracts oversee the distribution of the system’s computing resources, assigning confidential nodes for computing tasks.
• Supporting the necessary Service Level Agreement (SLA) and scalability is accomplished by grouping nodes into clusters and pools with automated Disaster Recovery (DR) mechanisms. 
• Additionally, the capability to establish geo-distributed clusters with efficient local gateways is provided. 
• The protocol also ensures secure storage through multiple network replication, encryption, and restricted access to trusted applications. 

With these features, the capacity to deploy a fault-tolerant game server, storage, and agents without dependence on specific hosts is achieved, embodying the finest decentralized cloud architecture available today. 

4. Ownership management via NFTs and smart contracts is central to the project. The entire process of AI agents’ ownership management and the orchestration of the marketplace where they are traded, is exclusively governed by the project’s smart contracts. 

Each agent consists of on-chain data — an NFT with its own wallet, and data in storage, including a model and game interaction history. Any modifications to the model are made within the chain of trust, beginning with the initial record. These changes are exclusively executed through a smart contract, with each alteration recorded in the agent’s track record and the blockchain. 

Listing of an agent on the marketplace and transferring ownership is seamlessly and securely conducted through a smart contract, ensuring the safety of all participants. 

To sum it up, the deployment of a seamless and secure confidential, verifiable, decentralized computing service, combined with state and ownership management through smart contracts, creates a competitive and fair environment for AI agents across diverse activities. 

Uncompromised competition translates into value, meaning that an AI agent’s success in gaming tasks transforms it into an asset that accrues both the player’s time and money, along with the diversity and uniqueness of gaming scenarios it has encountered. 

NFT integration grants the agent autonomy, meaning that it can be traded and rented. Moreover, the agent becomes capable of possessing its own assets and making decisions independently, acting autonomously without external influence. 

The Rest of Web3 Infrastructure: Polygon and Immutable 

Alongside the Super Protocol, the Veriplay team chose Polygon and Immutable X technologies, merging these three platforms to establish a resilient Web3-compliant ecosystem for training and dynamically tokenizing AI Agents. On a high level, each solution has the following functions: 

• Super Protocol provides the decentralized and confidential verifiable computing infrastructure necessary for widespread adoption of AI agents and dynamic NFTs in games. Leveraging TEE technology, Super ensures data and algorithms are safeguarded against

unauthorized access, manipulation, and attacks, crucial for establishing trusted and transparent web3 games. 

• Polygon offers a fast and scalable blockchain, delivering high performance and low transaction fees. This enables efficient management of AI agents and dynamic NFTs, ensuring a seamless and cost-friendly gaming experience. Moreover, Veriplay, with its focus on a multichain future, seamlessly integrates with other EVM-compatible networks through Polygon’s multichain framework. Additionally, Polygon’s compatibility with the Ethereum Virtual Machine grants Veriplay direct access to the vast capabilities of the Ethereum ecosystem. This ensures not only smooth scaling but also opens doors to a wider range of opportunities within the Web3 space. 
• Immutable X stands out as a premier NFT platform focused on gaming, offering scalability, low fees, and developer-friendly tools. These features simplify the integration of dynamic NFTs into games, requiring minimal cost and effort. 

Conclusion 

Super Protocol marks a new era in Web3 Games development, paving the way for innovative and immersive game worlds controlled by players and built on principles of trust and transparency. 

By enabling exclusive ownership of confidential, verifiable, and transferable AI Agents through dynamic NFTs, players can fully trust the authenticity of in-game assets. Moreover, investors in the open market can confidently invest in NFT assets backed by real AI models, sought after by players for gaming applications.

Author:  Sal Kimmich

In the realm of confidential computing, ensuring trust and security in computing environments is paramount. Attestation libraries and tools provide essential components to build systems that can produce and verify evidence of trustworthiness. This blog explores the concept of attestation in confidential computing and highlights two significant projects within the Confidential Computing Consortium (CCC): Veraison and SPDM Tools.

What is Attestation in Confidential Computing?

Attestation is the process by which the hardware provides evidence about itself and the software running under its protection. Any other party can use this evidence to evaluate the trustworthiness of the Trusted Execution Environment. This process is critical in confidential computing to establish and maintain trust in computing environments, ensuring that sensitive data and operations are protected from unauthorized access and tampering.

Key Components of Attestation

1. Evidence Generation:
   • The hardware (e.g., a device or CPU) generates evidence about its state, such as cryptographic measurements and signatures.
2. Evidence Verification:
   • The verifier evaluates the provided evidence against a set of policies or reference values to determine the entity’s trustworthiness.
3. Trust Anchors:
   • Cryptographic roots of trust (e.g., certificates) used to validate the identity.

Veraison: A Comprehensive Attestation Verification Service

Project Veraison builds software components to facilitate the creation of an Attestation Verification Service. Here’s how Veraison operates and its significance:

Overview

• Purpose: Veraison aims to simplify the development of attestation verification services by providing reusable software components. These components include verification and provisioning pipelines that can be extended with plugins to support specific attestation technologies.
• Flexibility: The project’s core components are designed to adapt to various deployment environments through abstractions, allowing for custom service creation without the need for extensive bespoke development.

Key Features

1. Verification Pipelines:
   • Core structures for verifying attestation evidence, ensuring that it meets established trust policies.
2. Provisioning Pipelines:
   • Components that manage the provisioning of data required for evidence appraisal, sourced from authoritative sources.
3. Extensibility:
   • Support for plugins allows the service to handle various attestation technologies, making it versatile and adaptable to different use cases.
4. Community and Collaboration:
   • Veraison is a collaborative project with active community involvement, including regular public meetings and contributions from multiple organizations.

Use Case: Veraison in Action

Veraison provides reference implementations to demonstrate integration principles, offering a convenient basis for developing substantive attestation verification services. These reference implementations showcase how the core components and plugins work together to create a robust verification system. 

Veraison also supports REST APIs to assist in end-to-end integration with attestation scemes, or can be used as verification components within a custom deployment. A great example of this is a key broker service, where successful attestation verification a key released to a Trusted Execution Environment. 

SPDM Tools: Enhancing Security with Attestation Protocols

SPDM (Security Protocol and Data Model) Tools offer libraries and utilities to implement the SPDM protocol, a standardized framework for secure communication and attestation between devices.

Overview

• Purpose: SPDM Tools provide essential functionality for implementing the SPDM protocol, ensuring secure communication and attestation across various platforms.
• Interoperability: The tools ensure interoperability between different devices and platforms, promoting a unified approach to security and attestation.

Key Features

1. Protocol Implementation:
   • Comprehensive support for the SPDM protocol, enabling secure communication and attestation across various platforms.
2. Utilities and Libraries:
   • A suite of tools and libraries that simplify the implementation and management of SPDM-based attestation solutions.
3. Standardization:
   • By adhering to the SPDM standard, the tools promote consistency and reliability in attestation processes across different devices and environments.

Use Case: SPDM Tools in Secure Device Communication

SPDM Tools can establish secure communication channels between devices, ensuring that each device can verify the trustworthiness of the other before exchanging sensitive information. This capability is crucial in scenarios such as building a trusted channel between an accelerator device like a GPU and a Confidential Virtual Machine (CVM)..

SPDM-RS: A Rust Implementation for SPDM Protocols

SPDM-RS is a project within the CCC that provides a Rust language implementation of the SPDM, IDE_KM, and TDISP protocols. These protocols facilitate direct device assignment for Trusted Execution Environment I/O (TEE-I/O) in Confidential Computing.

Key Features

1. SPDM Protocol Implementation:
   • Supports various SPDM requests and responses, including version negotiation, capability negotiation, algorithm negotiation, and more.
2. IDE_KM and TDISP Protocols:
   • Implements protocols for secure communication and device management, enhancing the trust boundary of Confidential Virtual Machines (CVMs).
3. Cryptographic Algorithm Support:
   • Includes support for cryptographic algorithms such as SHA-256/384/512, RSA, ECDSA, AES-GCM, and ChaCha20Poly1305.
4. Cross-Platform Support:
   • Designed to work across different platforms, ensuring broad applicability in various confidential computing scenarios.

Conclusion

Attestation libraries and tools are vital for ensuring the trustworthiness of confidential computing environments. Projects like Veraison and SPDM Tools within the Confidential Computing Consortium provide essential components for building robust attestation solutions. By leveraging these tools, developers can create systems that securely verify and manage trust, protecting sensitive data and operations from potential threats.

Author: Manu Fontaine

At Hushmesh, a U.S.-based Public Benefit cybersecurity startup, we see Confidential Computing as a foundational technology for all things digital, paving the way for an inherently secure and private Internet. Imagine a future where Confidential Computing underpins a “universal zero trust” model at the chip level, whereby privacy and security are built into our digital infrastructure instead of bolted on.

Traditionally, data security and privacy are bolted on after the fact with a patchwork of point-solutions on top of an insecure infrastructure. However, with Confidential Computing, these critical  elements can become inherent to the infrastructure, automated  directly at the chip level without human intervention.

Our vision at Hushmesh is to utilize Confidential Computing to build the Mesh, a global information space and infrastructure, like the Web, but with automated end-to-end cryptographic security and privacy built in for everything and everyone. As Hushmesh CEO Manu Fontaine puts it, “Confidential Computing is the necessary technology to deliver digital peace of mind at internet scale. The Mesh is the definitive solution to identity theft, data breaches, fakes, and fraud.”

The potential of Confidential Computing extends beyond what is currently imaginable. By embedding security into the very fabric of our digital infrastructure, we aim to eliminate the vulnerabilities that threaten our digital lives, and to move towards a future where trust is inherent, not an afterthought. The need for this transformation is urgent, and we must act now to secure our digital future.

Confidential Computing is not just a technological advancement but a paradigm shift. It challenges us to rethink how we approach security and privacy for the next phase of the digital age, pushing us towards an inherently secure and trustworthy Internet for everyone. At Hushmesh, we are excited to be at the forefront of this revolution, working towards a future where Confidential Computing is ubiquitous. Without Confidential Computing, universal zero trust is simply not possible.

Join us on this journey to redefine digital security and privacy with Confidential Computing. Together, we can shape a future where our digital lives are secure and private, where trust is inherent, not an afterthought. Your participation is crucial in this collective effort to make the Internet what we all need it to be.

Read The Case for Confidential report here.

Manu Fontaine is the Founder and CEO of Hushmesh, the public benefit corporation developing and operating the Mesh. You can think of the Mesh as a global information space, like the Web, but with universal zero trust built in. Secured by the Universal Name System (UNS) and Universal Certificate Authority (UCA), the Mesh delivers what the Web never could: the global assurance of provenance, integrity, authenticity, reputation, confidentiality, and privacy for all bits within it, be they code or data, at internet scale. The Mesh is the definitive solution to identity theft, data breaches, fakes, and fraud. Hushmesh is developing privacy-preserving wallet and verifier Mesh agents for DHS SVIP, alongside secure “meshaging” for the North Atlantic Treaty Organization Defence Innovation Accelerator for the North Atlantic (NATO DIANA) Secure Information Sharing Challenge. www.hushmesh.com