Sui Ecosystem AI Panorama: 2025 Edition

Core Conclusion: The integration of AI and the crypto economy has completed a paradigm shift from narrative-driven to utility-driven, with Sui accelerating its evolution into the core coordination layer of on-chain autonomous economies. Its Object-Centric Data Model, high-concurrency parallel execution engine, and natively integrated on-chain storage, fine-grained access control, and Verifiable Compute modules collectively reduce the execution friction of the Agent's entire lifecycle to a historical low—this aligns closely with the macro trend of privacy protection and data provenance becoming core constraints for institutional and consumer adoption. The maturity of stable payment channels and the positive cycle of the developer ecosystem together form the underlying support for the current transformation, while market attention has significantly shifted from conceptual hype to the delivery and validation of production-ready systems.

Against the backdrop of continuously rising apparent activity, actual progress shows significant non-uniform distribution characteristics: 1) The Agent framework, AI-native Launchpad, and deep stablecoin integration are driving real economic activities rapidly on-chain, forming a verifiable positive cash flow loop; 2) The on-chain overhead of verifiable computing, the island effect of decentralized identity systems, and the fragmentation of cross-chain data availability still constitute the main scaling bottlenecks at this stage.

The essential characteristics of this expansion cycle can be summarized as "selective distribution + systematic de-risking," rather than a mature stage where demand has been fully realized: the infrastructure supply curve significantly leads the complete ramp-up of the terminal demand curve, representing a typical late-cycle infrastructure pre-layout.

The current ecosystem does not show structural top signals; instead, it reflects typical characteristics of a late cycle shifting towards utility validation. The following three types of catalysts have entered an observable maturity window:

1. The acceleration of payment layer standardization, with continuous improvement in stablecoin circulation efficiency and compliance certainty, providing a certain settlement base for the Agent economy;

2. Lightweight verifiable computing solutions are gradually moving from research prototypes to mainnet deployment, with the cost of single proof expected to decrease to 1/5 to 1/8 of the original by Q1 2026;

3. The preliminary integration of cross-chain data markets and intent routing layers is forming a market-based pricing mechanism for decentralized data elements.

1. Introduction

The intersection of artificial intelligence (AI) and blockchain technology has become one of the most transformative dominant narratives in the entire crypto ecosystem by 2025, driven by the exponential enhancement of decentralized computing power, a comprehensive awakening of data sovereignty awareness, and the rapid prototyping and implementation of autonomous economic systems.

This report aims to provide a comprehensive and systematic assessment of the AI ecosystem on the Sui chain as of November 2025. As a high-performance Layer 1 public chain developed by Mysten Labs, Sui, with its Object-Centric Data Model, native parallel execution engine, and highly developer-friendly full-stack toolchain, has clearly positioned itself in the first tier of the deep integration of AI and blockchain, providing the most efficient underlying infrastructure for scalable AI Agents, privacy-preserving computing, and AI asset tokenization.

AI × Crypto Overall Trend: In 2025, this sector is experiencing explosive growth, with the total market value of AI-related crypto assets surpassing $36 billion by mid-year (an increase of over 13 times from $2.7 billion at the beginning of 2023) [Crypto Research]. Institutional funds are accelerating their entry, with major catalytic variables including: AI Agent-driven on-chain yield maximization mechanisms, on-chain agents with autonomous decision-making capabilities, privacy-preserving AI computing based on zero-knowledge proofs (ZKPs), decentralized data ownership models, and Agent-to-Agent (A2A) payment protocols.

Sui Chain Overall Trend: By mid-2025, Sui's total value locked (TVL) has surpassed $2 billion, firmly ranking 8th across all chains; monthly DEX trading volume remains stable at around $12 billion [Four Pillars, 4]. The Mysticeti V2 consensus protocol (which reduces finality delays by 35%), the upcoming mainnet Remora infinite horizontal scaling solution [Sui Foundation], combined with deep cooperation in the Asian region and ETF filing processes, collectively mark Sui's successful leap to a leading Layer 1.

Sui AI Ecosystem Trend: The AI track on the Sui chain has attracted over $150 million in funding (including Amazon AWS ecosystem funds and a milestone collaboration with Google Cloud on the Agentic Payments Protocol (AP2)) [CoinDesk]. Key technological breakthroughs include the Nautilus framework (which enables verifiable on-chain privacy AI computing) and deep integration with the Walrus decentralized storage network, providing high availability and low-cost off-chain data layers for AI models; combined with over 1.5 million AI computing nodes deployed under the DePIN path, Sui is being recognized by the market as the optimal candidate platform for the "global coordination layer" of AI Agents.

1.1 2025 AI × Crypto Trend: Institutional Convergence Towards Autonomous Economies

2025 marks a key maturity node in the AI × Crypto field, with institutional research reports recording a year-on-year growth rate of 28% in ecosystem financing, totaling $18 billion, of which the AI × Crypto sub-sector accounts for over 5% (with 48 rounds of financing totaling $201 million). This data highlights the investment attractiveness of the post-large language model (Post-LLM) era's integration paradigm [Root Data, 2--3]. This growth momentum primarily stems from the deep integration of cryptocurrency and AI, transitioning from early speculative phases to supporting essential applications such as DeFi, RWA, and stablecoins, with AI-driven startups accelerating their transformation from ideation to token generation events (TGEs) [Root Data, 3]. Broader macro indicators show that the global crypto market's total market value has reached $2.7 trillion, with Bitcoin surpassing Google to become the fifth-largest asset class, while the circulation scale of stablecoins has exceeded $240 billion (Citigroup predicts it will reach $1.6 trillion by 2030) [KPMG, 37--38]. The World Economic Forum estimates that by 2027, 10% of global GDP will be tokenized on the blockchain, with the scale of RWA potentially reaching $10 trillion in the next decade, and security tokens expected to exceed $30 trillion by 2030 [KPMG, 37--38].

These trends are reflected in four core pillars, each supported by clear institutional data, enabling scalable, trustless AI applications while effectively alleviating structural pain points such as centralization and data silos from the Web2 era.

1.1.1 On-chain AI Agents: Workflow Automation in Programmable Finance

On-chain AI Agents—autonomous entities executing multi-step tasks on the blockchain—have entered an accelerated iteration phase, with the deep convergence of generative AI (GenAI), Agentic AI, and blockchain giving rise to a new paradigm of "Programmable Finance," where Agents can dynamically interact with crypto protocols, achieving automation upgrades in DeFi [KPMG, 40]. Typical cases include the Mindshare Index based on the NEAR protocol, where the Agent deploys AI mechanisms to scan social sentiment indicators, execute cross-chain transactions, and autonomously rebalance portfolios [Mindshare Index on NEAR]; and Fetch.ai's Agent applications on DeFi platforms, which utilize machine learning algorithms to optimize asset swaps, identify arbitrage opportunities, and allocate liquidity [KPMG, 40]. Emerging DeFi lending Agents assess credit risk through wallet behavior analysis, supporting low collateral loans and introducing data-driven underwriting mechanisms [KPMG, 40].

In short, the core value of Agents lies in minimizing human involvement in market monitoring and smart contract triggering, as noted in the KPMG report: "AI Agents operating on decentralized networks can continuously monitor markets, execute trades, manage liquidity, and trigger smart contracts with almost no human intervention" [KPMG, 40]. It is expected that through deep integration with tokenized assets, they can unlock over $100 billion in efficiency increments for the DeFi ecosystem.

1.1.2 Privacy-Preserving AI: Ensuring Execution Security in Decentralized Environments

The core of privacy-preserving AI lies in preventing data leakage during the computation process through Trusted Execution Environments (TEE), with TEE-based architectures ensuring that model weights and input data remain invisible throughout [Four Pillars, 89]. McKinsey points out that current AI applications still face widespread risks of customer data and protected data leakage, directly driving a strong demand for sovereign clouds and localized deployments [McKinsey, 25, 43]. In the field of bioengineering, while AI can significantly accelerate research and development processes such as protein structure prediction, it also highlights the urgency for secure data processing mechanisms [McKinsey, 83].

As inference costs decrease annually by 9 to 900 times (with a median of about 50 times), privacy layers like TEE have become key thresholds for enterprise compliance adoption, effectively reducing the risk of intellectual property leakage during large model training [McKinsey, 23, 61].

1.1.3 AI Data Ownership and Monetization: Tokenizing Knowledge to Build Sovereign Economies

The trend of AI data ownership centers on structuring and tokenizing datasets for market circulation, with blockchain providing an immutable provenance mechanism [KPMG, 37]. Root Data shows that AI-driven RWA and tokenized assets have become financing hotspots, with this sub-sector accumulating $369 million in financing in 2025 (49 rounds) [Root Data, 2]. McKinsey cites examples in the field of earth observation, where AI can compress the climate analysis time of hyperspectral data from 57 minutes to mere seconds, giving rise to real-time tokenizable data assets [McKinsey, 91, 94].

Overall, the scale of tokenized assets is expected to reach $2 to $30 trillion by 2030, granting users true ownership through NFTs and DAOs. As the report states: "Decentralized identity and personal proof mechanisms based on blockchain may become the necessary infrastructure for distinguishing AI from humans in the future" [KPMG, 38, 41].

1.1.4 Agent-to-Agent Payments and Economic Coordination: The Blockchain Foundation for AI Transactions

A2A payments are becoming a core role of blockchain in AI economies, as described by KPMG: "AI-capable agents will transact online with humans and other agents, completing service settlements through payment channels such as stablecoins" [KPMG, 41]. Root Data indicates that AI-driven payment scenarios are particularly active in the RWA and stablecoin sectors, with stablecoin issuers accumulating $2.35 billion in financing (15 rounds) by 2025 [Root Data, 2]. McKinsey mentions JPMorgan's Kinexys blockchain foreign exchange settlement system, which has the capability to expand to micro-payments for Agents [McKinsey, 59]. With the circulation scale of stablecoins reaching $240 billion, A2A payments can achieve real-time settlements, as the report concludes: "Blockchain payment channels are likely to be the primary transaction method between AI Agents" [KPMG, 38, 41]. Citigroup predicts that the scale of stablecoins will reach $1.6 trillion by 2030.

1.2 2025 Sui Chain Core Trend: Scalability Drives Mainstream Adoption

The performance of Sui in 2025 fully reflects optimistic expectations at the institutional level, with its native token SUI increasing by 200% over the past year, significantly outperforming competitors like Ethereum and Solana, and the total market value of the entire network accounting for over 5% of Ethereum (with a relative valuation increase of 12.5 times) [21Shares].
Sui has successfully captured 15% of Ethereum's valuation share, with SUI price expected to reach $9.97, supported by technical indicators from the 50-day and 200-day moving averages [21Shares].

1.2.1 Technical Upgrades: Low Cost and High Throughput

Sui's transaction costs are approximately 3 times lower than Solana and about 150 times lower than Ethereum, ranking first in theoretical throughput among the top five smart contract platforms. The annualized revenue from network fees is expected to reach $15 million by 2025; the Ika protocol is set to launch on the mainnet in mid-July 2025, supporting native Bitcoin/XRP transactions without bridging. The Walrus decentralized storage protocol has been selected by Plume Network as the trusted layer for RWA data.

1.2.2 Major Partners and Ecosystem Dynamics

Sui has received strategic endorsements from a16z and Apollo Global; DeepBook, as Sui's native liquidity layer, has served Cetus and Bluefin (the two largest DEXs by trading volume on Sui over the past 30 days) [Grayscale Research; 21Shares]. Enterprise-grade stablecoins like First Digital's FDUSD and Ondo's USDY (accounting for over 10% of the $1 billion stablecoin supply) further strengthen ecosystem depth [Grayscale Research; 21Shares].

SuiPlay0X1's handheld gaming device has pre-sold over 9,000 units and sold out; Pebble City has generated over 14 million on-chain transactions since its launch in February 2025 [21Shares].

With developers and funds accelerating, combined with $1 billion in stablecoins poised for action, and with the ETF approval window gradually clarifying, Sui is on the eve of a large-scale influx of institutional funds, aptly described as "the bullet is in the chamber, just waiting for the regulatory trigger to be pulled."

1.3 2025 Sui AI Ecosystem Trend: Decentralized Infrastructure for Agentic Economies

The core focus of the Sui AI ecosystem in 2025 is on on-chain autonomy and privacy protection, with leading projects like Talus and Atoma fully leveraging Sui's object model for efficient Agent execution [Four Pillars, 88--89]. The deep integration with Walrus provides high-speed access to off-chain data for large models and embeddings while ensuring full auditability throughout the on-chain process [Four Pillars, 88].

1.3.1 Strategic Partnerships and Core Tracks

Key collaborations include: Talus partnering with Sui/Walrus to launch the Nexus framework, covering multi-scenario Agent applications in DeFi, gaming, and DAOs; Atoma using Sui smart contracts to record AI task metadata and settlement records; and InflectivAI collaborating with Walrus/Seal to build a tokenized data market under strategic control [Four Pillars, 88--90]. The focus is on decentralized AI markets, robotic data engines (OpenGraph), SocialFi Agents (VRAM.AI), and knowledge layers (InflectivAI) [Four Pillars, 86--90].

1.3.2 Financing and Ecosystem Highlights

Financing data for individual projects in 2025 has not yet been fully disclosed, but projects like FanTV have shown explosive growth: over 6 million active users, 75,000+ creators, and 1.2 million hours of monthly viewing time, all enabled by AI content tools [Four Pillars, 90]. The VRAM.AI testnet is distributing tVRAM tokens; the Genesis NFT on Walrus provides early rights [Four Pillars, 87].

1.3.3 Key Technological Breakthroughs

● Talus Nexus Framework: Encapsulates autonomous Agents as Sui objects, supporting plug-and-play transaction/workflow and revenue sharing [Four Pillars, 88].

● Atoma TEE Architecture: Samples consensus verification results and calls external large models (OpenAI, Mistral) via Sui, ensuring privacy execution [Four Pillars, 89].

● OpenGraph Robotic Engine: Manages robotic entities as Sui objects, zkLogin supports remote control, and on-chain rewards data set contributions [Four Pillars, 86].

Sui's "high throughput, object-centric data model" is naturally suited to the real-time settlement needs of AI, having substantially evolved into "the underlying foundation of decentralized autonomous AI systems" [Four Pillars, 89], resonating highly with the macro funding tide where global AI equity investment has reached $124.3 billion [McKinsey, 19].

2. AI × Web3 New Pattern

As of November 2025, the intersection of AI and Web3 has completely transitioned from a conceptual framework to a tangible infrastructure and economic model. Multiple institutions consistently predict that by 2030, AI-driven on-chain purchases will reach $30 trillion, while blockchain payment channels will become essential foundations for autonomous Agent transactions [a16z]. This section synthesizes the latest research reports from a16z, KPMG, EY, Architect Partners, and others, systematically breaking down the underlying logic of mutual empowerment, emerging technological paths, and business model loops.

Institutions have formed a strong consensus: AI is systematically breaking down the long-standing usability barriers of Web3; Web3, in turn, provides the most thorough antidote to the centralization risks inherent in AI. The hybrid system formed by their coexistence is expected to capture 10-15% of global GDP through asset tokenization by 2027 [KPMG, 37; EY].
The aforementioned progress is not an isolated phenomenon: a16z clearly states that crypto technology is the only viable solution to break the monopoly of OpenAI and Anthropic, which collectively account for 88% of generative AI revenue; EY emphasizes that blockchain is the foundational technology for resolving AI ethical disputes and establishing verifiable trust [a16z; EY; Architect Partners, 5--6].

The following will break down three core dimensions: collaborative mechanisms, emerging technology stacks, and business model loops, with specific cases and key indicators.

2.1 Mutual Empowerment: How AI and Web3 Achieve Innovative Crossovers

AI and Web3 have formed a typical symbiotic relationship: AI injects intelligence and scalability into decentralized systems, while Web3 provides transparency, ownership, and decentralization guarantees for the prevalent black-box and centralization tendencies in AI. Multiple leading institutional reports consistently highlight three core empowerment paths: verifiable trust and provenance, significantly enhanced usability and automation levels, and economic incentive mechanisms for data and computing power. These paths collectively give rise to a series of innovative crossover applications, such as verifiable content and adaptive organizations.

2.1.1 Main Paths of Mutual Empowerment

AI Empowers Web3: AI automates complex workflows on-chain, including dynamic pricing of computing resources and intelligent upgrades of consensus mechanisms. McKinsey's "2025 Technology Trends Outlook" shows that the demand for Agentic AI-related job postings surged by 985% in 2023-2024, continuing to grow in 2025, with AI Copilots significantly lowering the interaction barriers for Web3, simplifying navigation and personalizing experiences [McKinsey, 3; EY].

Web3 Empowers AI: Blockchain provides immutable provenance for AI training data, fundamentally addressing IP leakage and bias issues through tokenized datasets and federated learning solutions. a16z's latest report clearly states: "Crypto technology is the only antidote to counteract the centralizing forces of AI," especially in the context where Amazon, Microsoft, and Google collectively account for 63% of cloud computing power, with DePIN networks systematically achieving decentralization of computing power [a16z; Architect Partners, 3].

In terms of feasibility, Web3's ability to resolve the trust crisis in AI through verifiable ledgers like on-chain notarization, coupled with the annual decline in AI inference costs by 9 to 900 times (with a median of 50 times), has transitioned on-chain AI from "theoretically feasible" to "economically feasible" [EY; McKinsey, 23].

2.1.2 Innovative Crossover Applications and Benchmark Cases

Decentralized identity for distinguishing AI from humans: Worldcoin and Humanity Protocol have verified over 17 million people, providing "Proof of Human" to effectively combat bot accounts in the age of AI proliferation, enabling high-security interactions like anti-witch voting on-chain [a16z; EY].

AI Content Source Tracking: By using hash watermarks and on-chain notarization, AI-generated content is stamped with an immutable digital fingerprint, combined with multi-party computation (MPC) or zero-knowledge proofs for secure analysis and tracing [EY; a16z].

Tokenized model and dataset licensing: Platforms like Ocean Protocol support fractional ownership and automatic royalty distribution for AI models and datasets, opening up genuine decentralized monetization pathways amid high risks of AI IP leakage [a16z; Architect Partners, 4].

Key Insight: The aforementioned crossover innovations fundamentally mitigate the risk of model inversion attacks through immutable ledgers, while significantly catalyzing user growth in Web3: GenAI acts as the "intelligent co-pilot" of Web3, achieving personalized experiences, while DAOs evolve into new organizational forms with adaptability and intelligent decision-making capabilities under the support of GenAI [EY; McKinsey, 3].

2.2 Core Emerging Technologies Empowering AI × Web3

In 2025, protocol-level breakthroughs that truly bridge AI computing power demands with the trustless characteristics of Web3 are emerging one after another, with x402 becoming a benchmark infrastructure in the Agent payment track. The latest flagship reports from EY, a16z, and McKinsey cross-validate that the technological focus is highly concentrated on privacy protection standards and minimizing data exposure risks.

2.2.1 Key Technologies and Representative Implementations

x402 Protocol: By reapplying the long-dormant HTTP 402 "Payment Required" status code, it achieves seamless payments native to Web3, allowing AI Agents to complete micro-payments without intermediaries and supporting delayed settlements and batch billing models [a16z; EY]. a16z defines it as "the financial backbone of autonomous AI Agents." Against the backdrop of annual inference costs decreasing by 9 to 900 times (with a median of 50 times), combined with techniques like model distillation, x402 has demonstrated practical scalability to handle massive micro-transactions [McKinsey, 23].

Trusted Execution Environment (TEE): Achieving tamper-proof on-chain privacy AI inference at the hardware level, combined with on-chain notarization to form an undeniable audit loop for the computation process, has become the preferred solution for enterprise-level implementations [Architect Partners, 24; McKinsey, 25, 43].

Zero-Knowledge Machine Learning (ZK-ML): Completing AI computations off-chain, only bringing the results on-chain through zero-knowledge proofs for verifiability, achieving "data never leaves the local" federated learning, completely eliminating data exposure risks [Architect Partners, 5; a16z].
Agentic Framework and Multi-Party Secure Computation: A2A payments and multi-party computation (MPC) combined with ZKPs allow multiple parties to conduct joint analysis without exposing raw data, with the blockchain only responsible for result verification. This tech stack has been directly extended to Agent-level micro-payment scenarios, becoming the most promising infrastructure combination for 2025 [EY; a16z].

2.3 Major Business Models and Workflow Loops in AI × Web3

The path to the realization of AI × Web3 is highly focused on tokenized automation and data economies, with financing heat continuing to lead the entire sector: in 2024, AI + blockchain projects accounted for 8.5% of total crypto financing rounds and 10% of actual funds raised [Architect Partners, 2; Root Data, 2--3]. Almost all institutional reports consistently point out that revenue sharing achieved through staking and royalties has become the most mainstream and certain business model in this sector.

2.3.1 Core Models and Benchmark Cases

AI Agent Launchpad: Platforms like ChainGPT Pad and Virtuals Protocol complete Agent deployment and token issuance through one-stop IDO/IEO, achieving decentralized fundraising and automatic royalty sharing. In the Sui ecosystem, Surge officially launched in October 2025 and deeply integrated with the Cetus protocol (project application portal: surgeai.io), thoroughly streamlining the entire process of Sui's native AI Agent Launchpad, becoming one of the most significant infrastructure-level products in the Sui ecosystem in 2025 [Token Metrics; DappRadar; BNB Chain, 12; EY].

On-chain autonomous AI Agents: On-chain intelligent entities with complete execution capabilities, focusing on DeFi yield optimization (farming, arbitrage, etc.), with all operation records on an immutable ledger, generating transaction fees and staking rewards. A typical representative is Fetch.ai's Agent, which processes swap transaction volumes in the tens of billions annually, reducing human oversight costs by 70% [Architect Partners, 4; KPMG, 40].
Data labeling and monetization: By locking the labeling process on the blockchain, high-quality datasets are tokenized and continuously generate royalties, significantly accelerating AI training efficiency. Ocean Protocol predicts its data market will be valued at $40 billion by 2026; McKinsey cites an example where the analysis time of the same hyperspectral data is compressed from 57 minutes to a few seconds with AI assistance, directly giving rise to real-time tradable data assets [McKinsey, 91, 94; EY].

Revenue-sharing ecosystem: New DAOs empowered by GenAI automatically allocate incentives through smart contracts, promoting collaborative machine learning. The Bittensor network has directed 8.5% of its trading flow towards hybrid projects and introduced a second-round financing mechanism to ensure growth fairness, becoming a textbook case of successful revenue-sharing models in 2025 [a16z; Architect Partners, 5].

The current business model of the Web3 × AI sector has fundamentally shifted from speculation-driven to utility-value-driven. Architect Partners predicts that by 2033, the AI + blockchain market will reach $35 billion [Architect Partners, 2], but EY simultaneously warns that ethical risks such as large model hallucinations still objectively exist and must be systematically governed through verifiable provenance technologies like x402 [EY; Architect Partners, 5].

In summary, the cycle from Launchpad → on-chain Agents → data assets → revenue sharing has formed a complete and replicable business positive loop. With native infrastructures like Surge, Sui is accelerating its position as the optimal execution layer for this closed loop.

3. Sui's AI Infrastructure: Overview of AI Agent Development Technologies on Sui

Based on the symbiotic technological evolution of AI × Web3, Sui has established itself as a top-tier infrastructure layer for deploying autonomous AI Agents, with its native architectural innovations directly addressing core pain points such as decentralized computing, data sovereignty, and verifiable execution, providing systematic solutions.

As of November 2025, the Sui ecosystem fully supports the construction and operation of scalable AI Agents, with key frameworks including Nautilus for privacy-preserving off-chain computing, Walrus for decentralized data storage, Seal for programmable access control, and the overarching Sui AI Stack—a complete verifiable control plane that enables seamless integration of large-scale AI models and on-chain settlements.
Institutional research reports consistently position Sui as the "Global Coordination Layer for Agentic Economies," benefiting from the deployment of over 1.5 million AI nodes under the DePIN path, a technical upgrade of the Mysticeti consensus that reduces delays by up to 35%, and the recently launched Verifiable AI Control Plane—ensuring accountability mechanisms across data, models, and actions [Four Pillars, 88--89; Sui Blog, Nov. 14, 2025; Messari, Q2 2025, Usage sec.].

This section will provide a fine-grained technical analysis of Sui's AI infrastructure, covering the development process and comparative advantages relative to traditional platforms like Ethereum and Solana.

3.1 Technical Overview: Core Components of AI Agent Development on Sui

The core architecture of Sui's infrastructure is built around the Object-Centric Data Model and parallel execution engine, which collaboratively achieve efficient, high-throughput processing of AI workloads, including Agentic decision-making, federated learning, and tokenized data markets. Unlike traditional account models, Sui treats digital assets as independent "objects" with unique IDs, requiring transactions to reference and modify specific objects without global state lock contention. This model relies on the Move programming language, supporting the deployment of AI Agents as programmable Sui objects—these self-contained entities possess autonomous execution, state management, and interaction capabilities with external oracles and large language models (LLMs) [Sui Docs; Grayscale Research, sec. 1].

Today, the Sui AI Stack, composed of Walrus, Seal, Nautilus, and the Sui core, has formed a complete verifiable control plane from data ingestion to execution proof [Sui Blog, Oct. 28 & Nov. 14, 2025].

Core Technology Components Explained

Object-Centric Architecture: The Native Foundation of Agent Autonomy - AI Agents on Sui are instantiated directly as objects defined by Move, inherently possessing key (on-chain unique identification and addressing) and store (mutable state storage) capabilities. Developers can precisely constrain Agent behavior through the entry function of Move modules, eliminating unauthorized state changes—this is crucial for secure and trustworthy Agent-to-Agent (A2A) interactions.

Typical Case: A DeFi yield optimization Agent can be designed as a singleton object, setting persistent parameters (like portfolio weights) through a module initialization function, while each asset holding is supported as independent sub-objects for parallel updates, completely eliminating resource contention in traditional global storage models. In practical multi-step workflows, the object's "borrowing" mechanism (temporary reference without transferring ownership) can reduce gas costs by up to 50%. In multi-Agent collaborative scenarios, shared state updates must undergo cryptographic verification to ensure consistency and security [Sui Docs; Messari Q2 2025 Primer; Sui Blog Nov. 14, 2025].

Parallel Execution Engine: The Key to Scaling AI Inference - Sui employs dependency-aware scheduling based on object IDs, allowing inherently independent transactions to run in parallel, completely freeing itself from sequential execution bottlenecks. This is particularly suitable for AI Agents requiring real-time inference, such as social data sentiment analysis and DEX liquidity pool arbitrage detection in high-frequency scenarios. Verification nodes only need to horizontally scale to achieve nearly linear TPS growth (8 times hardware ≈ 8 times throughput), with simple transfer "fast paths" maintaining latency stability in sub-second ranges. Combined with Mysticeti consensus, the finality of blocks is compressed to about 400ms (a 35% improvement over previous generations), enabling Sui AI to perfectly adapt to time-sensitive Agent-level micro-payments. The latest cases include the on-chain native base model launched by Zark Lab, which has achieved real-time on-chain transaction risk detection [Grayscale Research sec.1; Four Pillars 88; Messari Q2 2025 Infrastructure].

Nautilus Framework: Verifiable + Privacy-Preserving Off-Chain Computing - Nautilus extends Sui's on-chain primitives to verifiable execution off-chain, relying on TEE hardware enclaves like AWS Nitro Enclaves. Developers can offload high-computational tasks (such as private model training and ZK-ML proof generation) to TEE while retaining on-chain result verifiability.

Standard Workflow:

1. Encrypted input submitted as a Sui object →
2. Computation completed within the TEE enclave →
3. Generates on-chain verifiable encrypted proof →
4. Outputs settled on-chain as a new object

This mechanism completely avoids data leakage risks in federated learning scenarios and provides one-click integration with external LLMs like OpenAI and Mistral.

Nautilus was launched on the mainnet on June 5, 2025, and has since driven verifiable AI inference for Bluefin Pro, federated learning solutions, and several champion projects from the Sui Overflow 2025 hackathon (599 projects, including AI-focused categories) [Four Pillars 89; Messari Q2 2025 Infrastructure; Sui Blog Nov. 14, 2025].

Walrus: Decentralized AI Data Storage and Retrieval - Walrus, as Sui's programmable blob storage protocol, launched on the mainnet on March 27, 2025, specifically addresses the storage needs of AI for massive datasets, supporting the storage of large-scale datasets (such as embeddings or hyperspectral training corpora) off-chain while retaining on-chain metadata and access control.

Through its unique erasure coding technology, Walrus achieves 80% higher storage efficiency than competitors like Filecoin; permission management is jointly completed by Sui objects combined with zkLogin (an OAuth-based authentication method) and gas-sponsored transactions (user zero gas operations). For AI Agents, this means they can dynamically and securely pull data from off-chain (e.g., Talus Nexus Agent calling historical market data for on-chain decision-making) without any centralization risks, and it naturally supports a tokenized data market with immutable provenance.

As of June 2025, Walrus's total storage capacity has surpassed 1 PB, having integrated with decentralized training projects like io.net and OpenGradient [Grayscale Research, sec. 3; Messari, Q2 2025, Infrastructure sec.; Sui Foundation, Mar. 27, 2025; Sui Blog, Oct. 28, 2025].

Seal: Programmable Access Control and Other Key Enabling Components - Seal provides programmable decryption strategies for encrypted data on Walrus, such as token-gated access based on requester identity, purpose, and timeliness, ensuring compliant data sharing in collaborative AI scenarios. zkLogin allows AI Agents to authenticate users directly through Web2 accounts, significantly lowering the entry barriers for hybrid AI-Web3 applications; gas-sponsored transactions accounted for 25% of all network transactions by Q2 2025. Atoma's TEE node network and the latest accelerator projects like Bluwhale further deepen integration with Sui, achieving decentralized inference with verifiable results but zero data exposure through sampling consensus [Messari, Q2 2025, Usage sec.; Four Pillars, 89; Sui Blog, Nov. 14, 2025].
These components collectively form a highly collaborative full-stack for AI Agent development, emphasizing auditability and scalability. Institutions expect that under the catalysis of a $150 million ecosystem fund and innovations driven by hackathons (such as the crypto data market), the deployment of AI nodes in Sui will triple by the end of 2025 [Four Pillars, 88; Messari, Q2 2025, Developer sec.].

3.2 Technical Details and Development Process: AI Agent Development on Sui vs Other Platforms

Developing AI Agents on Sui adopts a highly streamlined object-oriented process, contrasting sharply with the typical sequential execution and account-centric paradigms of Ethereum and Solana, significantly reducing the complexity and runtime costs of Agentic applications.

Sui Native Development Process

1. Write modules using the Move language to define the Agent structure (e.g., struct Agent has key, store { id: UID, state: vector, params: MyParams });
2. Publish with one-click via Sui CLI or SDK;
3. Complete initialization through the entry function.

Subsequent interactions of the Agent are all based on programmable transaction blocks (PTB), allowing for parallel scheduling by declaring object inputs in advance. For example, an Agent executing yield farming can, in a single atomic transaction:
● Reference the liquidity pool object → Call Nautilus to complete off-chain risk assessment in the TEE enclave → Pull encrypted data controlled by Seal's strategy from Walrus → Complete settlement through A2A payments, with finality throughout at approximately 400ms.

● zkLogin + gas-sponsored transactions completely eliminate user-side friction, while the verifiable receipts provided by Nautilus and Sui ensure that every step (from data queries to computation proofs) has a complete auditable provenance [Sui Docs; Grayscale Research sec.1; Sui Blog Nov. 14, 2025].

Verifiable AI Control Plane: Typical Multi-Agent Process

(1) Agent requests data from Walrus using a verifiable ID →
(2) Seal decides whether to decrypt based on purpose-binding strategies →
(3) Nautilus executes inference or task orchestration in a confidential environment and issues proof →
(4) Sui is responsible for recording events and receipts on-chain, achieving full-link auditing under privacy protection conditions.

This process can support advanced applications like robotic fleets and network-level Agents, with performance scaling linearly with size [Sui Blog Nov. 14 & Oct. 28, 2025].

Comparison with Ethereum (Solidity/EVM): The Ethereum account model requires global state read/write, leading to sequential execution for multi-object operations and intense gas competition. Equivalent Agents must obtain off-chain data through Solidity contracts and external oracles like Chainlink, inherently lacking parallel capabilities, with average costs around $0.87 (100 times that of Sui) and finality of 12-15 seconds. Verifiable computing relies on external Layer 2 solutions like Polygon zkEVM, leading to fragmented tech stacks and complex provenance tracking; Sui's native integration of Nautilus and Seal achieves up to 150 times cost savings on end-to-end process costs and enforces strategy compliance [Grayscale Research Exhibit 2; Messari Q2 2025 Primer].

Comparison with Solana (Rust/Sealevel): While Solana achieves parallelism through Sealevel, the granularity of parallelism is limited by account locks, leading to severe competition in shared state AI scenarios (e.g., multiple Agents trading simultaneously). Development requires using Rust and BPF loaders, with cost fluctuations reaching up to 3 times that of Sui, and multiple congestion events in 2024 have compromised reliability. Sui's object model completely decouples competition, providing deterministic pricing and horizontal scalability; the data layer lacks native solutions at the level of Walrus and Seal, relying instead on external storage like IPFS, resulting in higher retrieval latency and weaker strategy control [Grayscale Research sec.1; Messari Q2 2025 Primer].

3.3 Relative Advantages of Sui as AI Infrastructure

Sui's advantages stem from its architecture designed specifically for low-latency, highly composable AI, surpassing Ethereum's "mature but obviously bottlenecked" old-school ecosystem and outperforming Solana's "extreme speed but cost volatility" radical model.

From an institutional data perspective, as of Q2 2025, Sui's TVL reached $1.76 billion (up 44.3% quarter-on-quarter, ranking 8th across all networks), with AI-native projects like Zark Lab and Bluwhale contributing 29 million transactions and daily DEX peak trading volumes of $367.9 million; while Ethereum's TVL is as high as $120 billion, the development costs for AI Agents remain about 40% higher [Messari Q2 2025 Usage; Grayscale Research sec.3].

Sui's highly integrated tech stack not only accelerates AI Agent deployment but also drives the formation of fair and sovereign new economic systems from the ground up. Institutions expect that by the end of 2025, Sui's annualized fee revenue will reach $15 million, securing its optimal Layer 1 position in the $30 trillion Agent economy wave, while the verifiable AI control plane as a core competitive advantage marks a critical evolution in the accountable AI design paradigm [Grayscale Research sec.3; a16z 2025; Sui Blog Nov. 14, 2025].

4. Overview of the Sui AI Ecosystem

Leveraging the hard power of Sui's underlying tech stack, the Sui AI ecosystem has rapidly evolved into a leading hub for decentralized intelligence by 2025. In just Q1, the ecosystem attracted over $21 million in funding, primarily from AI-specific grants from institutions like AWS and deep collaborations with Google Cloud on the Agentic Payments Protocol (AP2); over 1.5 million nodes have been deployed under the DePIN path, forming a highly modular AI computing and data network [Launchy, May 16, 2025; BeInCrypto, Jan 2, 2025; Four Pillars, 88].

As of November 2025, the ecosystem has entered a new acceleration phase:

● The compliant stablecoin USDsui launched on November 12
● The Beep protocol (AI-native smart wallet framework) was officially released

Together, these two initiatives directly embed the automation capabilities driven by large models into high-frequency DeFi and payment tracks, with the total stablecoin transfer scale exceeding $412 billion in just two months from August to September [Sui Blog Nov 12, 2025; CoinDesk Nov 12, 2025; CryptoBriefing Nov 12, 2025].

This section will systematically break down recent key developments, leading projects, the synergistic amplification effects across the entire chain ecosystem, real revenue contributions, and institutional funding attention, comprehensively presenting Sui's strategic value and hard fundamentals as the optimal coordination layer for the AI economy.

4.1 Vision for the Sui AI Ecosystem - Recent Updates, New Ecosystem Support, and Technology Upgrades, Overview of Mainstream Projects

In the second half of 2025, the Sui AI ecosystem has shown significant acceleration, with ecosystem support continuously expanding through special funds, hackathons, and strategic collaborations, focusing on building verifiable AI Agents and data sovereignty. The Sui Foundation's $150 million fund plan (including AWS ecosystem funds and collaboration with Google Cloud on the Agentic Payments Protocol (AP2)) has directly funded multiple AI-oriented projects, with institutional research reports consistently indicating that this will further drive DePIN growth [Four Pillars, 88].

The Sui Overflow 2025 hackathon concluded successfully in June, receiving 599 project submissions covering nine major tracks, including AI, DeFi, and privacy, ultimately producing 36 award-winning projects. Among the 36 winners and 599 participants, the AI track ranked as the second most popular, thanks to Sui's top performance in the AI field. Typical representatives include advanced verifiable AI tools and federated learning prototypes [Sui Blog, Jun 30, 2025; Binance Square, Jun 30, 2025].

Technology upgrades include the launch of the Mysticeti v2 consensus protocol in November 2025, achieving a 35% reduction in delays and supporting real-time AI inference; simultaneously, the Nautilus framework was launched on the mainnet in June, providing TEE-based verifiable computing [CoinMarketCap, Nov 13, 2025; Sui Docs, Nautilus].

New ecosystem support includes the USDsui stablecoin launched on November 12, 2025, through the Bridge platform, further enhancing AI micro-payment capabilities, supporting U.S. compliance standards and integration with Stripe; during the same period, the total stablecoin transfer scale processed by Sui reached $412 billion [CoinDesk, Nov 12, 2025; CryptoBriefing, Nov 12, 2025; Cryptopolitan, Nov 12, 2025].

High-performing benchmark projects (based on financing scale, testnet activity, and integration depth) include:
Surge: Sui's first native AI Agent Launchpad, launched on October 28, 2025, supports community-driven IDOs and deeply integrates with Cetus DEX to fund AI projects; Adapt was selected from over 70 applications as the first multi-Agent quantitative trading protocol, unlocking early capital for Agentic startups [Bitget, Oct 28, 2025; Crypto News, Oct 29, 2025].

Talus Network: The Nexus autonomous AI Agent framework attracts deep community participation through the incentivized testnet (concluding on November 15, 2025) and the Talus Hub task system; it secured over $10 million led by Polychain Capital to support verifiable on-chain decision-making in DeFi and prediction markets [Decrypt, Sep 29, 2025].

VRAM.AI: The SocialFi attribute AI Agent Launchpad testnet is distributing tVRAM tokens, supporting monetization through NFTs, and achieving key milestones in community governance [Four Pillars, 87].

Beep Protocol: As Sui's first Agentic wallet framework, it launched on November 6, 2025, supporting LLM-driven DeFi automation, with early integrations demonstrating significant yield optimization effects [Sui Blog, Nov 6, 2025].

DeAgent AI (AIA): An AI Agent infrastructure across Sui, BSC, and BTC ecosystems, providing trustless autonomous decision-making capabilities; the token surged by 150% in early November, integrating AI oracles for prediction markets and DeFi protocols [Bitget, Nov 4, 2025; RootData, 2025; Medium, Nov 12, 2025].

These projects collectively drive ecosystem expansion, with Sui's TVL reaching $1.76 billion in Q2 2025 (up 44% quarter-on-quarter), significantly boosting transaction volumes through AI integration [Messari, Q2 2025, Usage sec.].

4.2 AI Projects on Sui: Roles, Capabilities, Classification, and Application Scenarios

Sui AI projects cover various sub-tracks, including Agentic Frameworks, Privacy Compute, Data Marketplaces, and Launchpads, each deeply relying on the Sui AI Stack (Walrus, Seal, Nautilus) to achieve verifiable and highly scalable operations. Core roles focus on autonomy (e.g., decision execution), privacy protection (e.g., secure inference), and monetization mechanisms (e.g., tokenized assets), with typical application scenarios including DeFi, gaming, SocialFi, and RWA.

Institutional reports indicate that total crypto financing reached $8.21 billion in Q3 2025, marking a quarterly high since Q1 2023, with the AI × Crypto crossover field being highly sought after—AI startups accounted for 51% of the total venture capital in the year [Messari, Q3 2025; Economic Times, Oct 22, 2025]. In this process, the Sui ecosystem attracted over $21 million in funding in just Q1, primarily directed towards core AI-driven projects like Agentic frameworks and DePIN nodes [Launchy, May 16, 2025].

4.3 Empowerment of the Sui AI Ecosystem: Leverage Effects on Non-AI Tracks

Sui AI projects are no longer isolated tracks but have become an enabling layer for the entire ecosystem, systematically enhancing the functional boundaries and scalability of traditional sectors like DeFi, RWA, and gaming through intelligent automation, secure data oracles, and verifiable computing. Thanks to Sui's native high composability of the object-centric model, AI-driven insights and execution capabilities can be seamlessly embedded into existing protocols without introducing any centralization risks.

Typical Empowerment Cases

Talus Network (Nexus Framework): Encapsulates autonomous AI Agents as Sui objects, achieving real-time automation on-chain; deeply integrates with DeFi protocols for yield optimization and arbitrage, while securely pulling off-chain data through Walrus, fully automating previously highly manual multi-step processes, achieving scalable implementation in lending and trading scenarios [Messari, Q1 2025].

Beep Protocol (Agentic Wallet Framework): Allows large models to directly drive programmable transactions across DeFi venues, significantly enhancing liquidity provision and yield strategy efficiency on platforms like Cetus, thanks to Sui's sub-second finality and extremely low settlement costs [Sui Blog, Nov 6, 2025; CoinTrust, Nov 10, 2025].

DeAgent AI (Oracle Infrastructure): Provides trustless AI-driven predictions for DeFi risk assessments; through on-chain data analysis, it enables low-collateral lending for protocols like Navi Protocol (credit scoring + market predictions), significantly reducing operational costs in dynamic environments.

RWA Track: InflectivAI's tokenized datasets combined with Walrus provide provenance-verifiable data layers, significantly accelerating the compliance processes for asset tokenization and cross-chain interoperability [Four Pillars, 90].

Gaming/DePIN Track: The OpenGraph robotic engine manages on-chain data set rewards for chain games like Pebble City, contributing over 14 million real transactions since its launch in February 2025, creating strong positive feedback through remote control and asset management.

Quantitative Results: The aforementioned synergistic effects have driven Sui's DeFi TVL to soar by 200% year-on-year, with institutions estimating that through AI-enabled programmable workflows, over $100 billion in efficiency dividends will be unlocked in the tokenized asset and automated finance sectors over the next 3-5 years [RootData, 3; KPMG, 40]. Sui AI is no longer just "the icing on the cake," but has become the underlying engine and core leverage for growth across the entire ecosystem.

4.4 Revenue Mechanisms and Growth Catalysts of Sui AI Projects

Sui AI projects achieve multi-dimensional monetization through tokenized incentives, profit-sharing mechanisms, and operational efficiency improvements, with institutions predicting that by the end of 2025, the annualized fee revenue across the network will reach $15 million, corresponding to the macro backdrop of the AI Crypto sector's total market value exceeding $36 billion.

Core Project Revenue Path Breakdown

Talus Network (Nexus Framework): Users stake $US to gain Agent verifiability and workflow audit services, with node operators and users sharing the revenues generated from DeFi and prediction market automation, forming a sustainable incentive loop.

Surge (AI Agent Launchpad): Since its launch on October 28, 2025, it has provided seed round financing channels for Agentic projects through community-driven IDOs and deep integration with Cetus DEX, establishing a shared treasury mechanism, allowing individual investors to participate in long-term AI projects [Bitget Oct 28, 2025].

VRAM.AI (SocialFi Launchpad): Drives user participation through dual-layer rewards of Token + NFT, linking with content platforms like FanTV (with over 1.2 million hours of monthly viewing time) to monetize AI content tools, with community governance and creator incentives forming strong network effects [Four Pillars 90].

Beep Protocol + USDsui: Together, they bridge A2A micro-payments and DeFi automation, supporting zero-fee USDC settlements, directly contributing to the $412 billion stablecoin transfer scale in August and September 2025 [Sui Blog Nov 6, 2025; OpenExO Nov 13, 2025].
Atoma Network (TEE Inference Nodes): Supports staking computing power to earn decentralized AI processing rewards, with its risk assessment and settlement functions contributing approximately 48% of Sui's monthly average DEX trading volume of $7.8 billion [Medium Jul 31, 2025].

Ecosystem Growth Evidence: In the first half of 2025, Sui's monthly active users exceeded 40 million, with daily active addresses around 490,000, significantly improving user retention through AI-driven DeFi and SocialFi applications [Grayscale Research sec.1; Four Pillars page 4]. Institutions further predict that by 2033, the total value of the AI + blockchain market will reach $35 billion, with Sui's DePIN expansion in verifiable computing and data sovereignty showing clear first-mover advantages [Architect Partners 2]. To date, Sui AI projects have fully transitioned from "technical concepts" to the mature business stage of "verifiable cash flows + sustainable incentives," becoming the dual-driving engine for revenue and user growth across the network.

4.5 Institutional Participation: Venture Capital and Strategic Investment in the Sui AI Ecosystem

In 2025, the Sui AI ecosystem attracted significant attention from institutional funds, with frequent strategic financing rounds and collaboration agreements, fully validating its strategic potential as a leading decentralized AI platform.

Polychain Capital led a $10 million financing round for Talus Labs in September 2025, focusing on supporting predictive AI and on-chain Agent innovations, with co-investors including the Sui Foundation and Walrus [Decrypt, Sep 29, 2025].

Momentum Ventures strategically invested in DeAgent AI on August 14, 2025, focusing on its cross-chain oracle infrastructure and emphasizing trustless Agent coordination capabilities [Gate.io, Aug 13, 2025].

Mill City Ventures committed $450 million to inject into Sui's ecosystem treasury tools in July 2025, including multiple AI-driven projects [Business Wire, Jul 28, 2025].

SUI Group Holdings announced its Q3 financial report on November 12, 2025, emphasizing the path of blockchain and AI integration, with total gross revenue of $2.6 million (a threefold year-on-year increase), focusing on digital asset strategies [Business Wire, Nov 12, 2025].

Overall, the Sui AI ecosystem not only amplifies on-chain autonomy but also positions the entire network as the core bearer of the $30 trillion tokenized GDP by 2030, with the deep integration of institutional capital and technological strength becoming a definitive trend [a16z, 2025; KPMG, 37].

5. Future Outlook for the Sui AI Ecosystem

As Sui gradually solidifies its leading position as a high-throughput Layer 1 public chain in the deep integration of AI × blockchain, the development trajectory of its AI ecosystem in the second half of 2025 and beyond will heavily rely on scalable expansion, interoperability optimization, and regulatory compliance alignment, while fully leveraging the inherent advantages of the object-centric architecture in Agentic applications.

Institutional forecasts indicate that the global AI × Crypto sector is expected to exceed $35 billion by 2033, with DePIN and Agentic frameworks driving 15-20% annual compound growth; with over 1.5 million nodes deployed and a $150 million ecosystem fund, Sui is in an excellent window to capture excess shares through verifiable computing innovations like Nautilus and Walrus [Architect Partners, 2; Four Pillars, 88].

This section will examine the latest developments in the competitive ecosystem, outline Sui's exclusive development path, and identify key shortcomings to provide a clear roadmap for sustainable expansion.

5.1 Competitive Ecosystem Progress Comparison: Next Steps for Similar Platforms

Competitive Layer 1 and Layer 2 ecosystems are accelerating AI integration to seize the $30 trillion Agent economy pie by 2030, focusing on DePIN computing power supply, multi-chain oracle data sovereignty guarantees, and hybrid frameworks for A2A coordination [a16z, 2025; Messari, Q3 2025].

Ethereum leads the influx of developers in 2025 with an additional 16,000 developers, followed closely by Solana with 11,500 new developers, while emerging L1s like Aptos and Near emphasize scalability based on Move and chain abstraction technologies [Yahoo Finance, Oct 16, 2025]. The following table summarizes the next steps, core metrics, and insights for Sui from key platforms like Ethereum, Solana, Base, BNB Chain, Hyperliquid, Aptos, Near, Avalanche, and Cosmos:

In this intense race, the Sui AI ecosystem stands out with the privacy protection and low-latency verifiable computing achieved through Nautilus TEE: inference speeds are 35% faster than Solana's heavily GPU-dependent Render network and significantly better than the latency of Ethereum's ZK-Rollup; simultaneously, the $150 million fund has catalyzed the deployment of 1.5 million DePIN nodes, with deployment speeds far exceeding Aptos's approximately 20 AI dApps, creating a clear generational gap in real-time A2A coordination scenarios [Four Pillars 88; Stakin Aug 13, 2025].

Advantages (Pros): The object-centric model brings 150 times the cost efficiency of Ethereum's fragmented ZK tools and surpasses Solana's prone-to-congestion account lock model, establishing segmented barriers in DeFi automation and SocialFi monetization—competitors like Base still lag significantly in native privacy layers.

Shortcomings (Cons): The ecosystem's maturity still has gaps: Sui AI projects number around 10-15, lagging behind Ethereum's 150 federated learning projects and Near's 50 abstract tool teams, posing risks of data silos, with cross-chain RWA scalability currently inferior to the liquidity pool depth of Aptos + Aave [DL News Aug 26, 2025; Panews Lab 2025].

5.2 Sui AI Evolutionary Strategic Path

The Sui AI roadmap has clearly defined three main lines: scaling Agents, multi-chain interoperability, and DePIN maturation, aligning closely with industry trends (Agentic job demand surged by 985% year-on-year, with AI equity investment reaching $1.1 billion) [McKinsey 3].

Path One: Cross-chain AI Data Market - Expand Walrus + Seal into a cross-chain tokenized data foundation, achieving federated learning with Solana and Ethereum; pilot retail-level AI investment channels through the Surge Launchpad in Q4 2025, tapping into the massive $1.5 trillion global AI spending market in 2025, capturing a share of the $100 billion AI data market by 2030, significantly lowering non-institutional participation barriers through community IDOs [Gartner Sep 17, 2025; Medium Jun 9, 2025].

Path Two: Advanced Agent Framework - Deeply integrate Nautilus with external large models like Mistral, embedding ZK-ML proofs directly into A2A payment protocols to achieve fully privacy-protected, verifiable multi-step autonomous transactions (such as automatic yield rebalancing across pools); the 2026 roadmap plans to increase node scale from 1.5 million to 10 million through layered DePIN incentives (robotic data staking rewards + SocialFi participation bounties), supporting ultra-large-scale applications like game DAOs and sentiment-driven creator economies, expected to drive a threefold growth in ecosystem TVL [MEVX Sep 14, 2025; Gate.io Apr 22, 2025].

5.3 Identified Shortcomings and Native Mitigation Strategies

Despite strong momentum, institutional reports in the first three quarters of 2025 still highlight three major hard issues: verification overhead, identity silos, and regulatory uncertainty [Launchy May 16, 2025; Messari Q2 2025].

Technical Bottlenecks and Iteration Cycles: ZK proof verification imposes additional computational burdens on AI Agents, slightly slower than Ethereum's mature ZK-ML standards; response strategy: internally strengthen Nautilus's TEE-ZK hybrid path to accelerate on-chain proof, while Surge launches a no-code Launchpad to abstract Move complexity, significantly shortening Agent construction cycles [AI Journal Sep 15, 2025; Gate.io Apr 16, 2025].
Identity Fragmentation and Witch Attacks: Cross-chain silos exacerbate risks for AI oracle robots, currently lacking Near-level anti-witch primitives; response strategy: embed DeAgent and other AI oracles into zkLogin to achieve dynamic "Proof of Human," allowing A2A interactions without external dependencies [Sui Blog Oct 10, 2023; Delphi Digital Aug 8, 2025].

Regulatory Uncertainty and Liquidity Gaps: U.S. crypto-AI regulations are still evolving, with RWA pilot volatility exceeding the stablecoin compliance trajectory of BNB Chain; response strategy: utilize AP2's built-in micro-payment proofs + $150 million fund to incubate self-auditing RWA modules, embedding Seal strategies into compliant data flows to smooth out impacts [State Street Mar 2025; Quinn Emanuel Aug 18, 2025].

In summary, the Sui AI ecosystem has transitioned from "technological leadership" to the "strategic closed-loop" stage. As long as the aforementioned paths are accurately followed to address shortcomings, it is highly likely that Sui will become the undisputed leader in the AI × Crypto sector by 2026.

Sui - The Vanguard of Decentralized AI Economies

In reviewing this report's systematic analysis of the deep integration of AI and blockchain, Sui has evolved from merely being a high-performance Layer 1 to clearly becoming the core coordination layer of the next generation of autonomous, sovereign digital economies.

From the overall explosion of the AI × Crypto sector in 2025 (with ecosystem financing reaching $18 billion and related token total market value at $36 billion) to Sui's unmatched scalability demonstrated through its object-centric parallel execution and the verifiable computing system built with Nautilus and Walrus, a technological moat has already formed [Four Pillars 4; Grayscale Research sec.1].

Within the ecosystem, star projects like Talus Nexus (benchmark for Agentic automation) and Surge (democratized Launchpad) are genuinely empowering DeFi, RWA, and SocialFi through privacy-preserving inference and tokenized incentives, driving rapid growth in TVL and attracting over $500 million in net institutional capital inflow [Messari Q3 2025 DeFi sec.; RootData 3].

Looking ahead, Sui has anchored three definitive strategic mainlines:

● Cross-chain AI Data Market

● Advanced ZK-ML Deep Integration

● Regulatory-Friendly Tool System

These three strategic mainlines position it to potentially surpass Solana's high TPS Agents and Ethereum's ZK standards by 2026, capturing 15% of Ethereum's valuation through AI ETFs and DePIN expansion [21Shares Q2 2025; a16z 2025]. Of course, thoroughly addressing identity fragmentation and the speed of development iteration will remain decisive variables in seizing the $30 trillion tokenized GDP dividend by 2030 [KPMG 37].

Ultimately, the Sui AI ecosystem has long since crossed the speculative phase and is forging a resilient underlying infrastructure: here, Agents autonomously optimize yields, tokenize knowledge, and coordinate economies without intermediaries. With institutional capital accelerating convergence (Q3 venture capital reaching $8.21 billion + Google AP2-level collaborations), Sui is ready to redefine trustless intelligence, empowering developers, users, and institutions to collaboratively build a programmable future for finance and beyond [Messari Q3 2025].

In this epic fusion of AI and blockchain, Sui is not just a participant but a true leader.

It is leading us toward a new world driven by intelligent agents and explosive value release.