Source: Cointelegraph Original: "{title}"

Opinion: Felix Xu, Co-founder of ARPA Network and Bella Protocol

Since 2024, artificial intelligence (AI) has been a mainstream topic, but users and companies still struggle to fully trust it. Whether it’s financial, personal data, or medical decisions, there remains a high level of hesitation regarding the reliability and integrity of AI.

This growing trust deficit in AI has now become one of the biggest barriers to widespread adoption. Decentralized, privacy-preserving technologies are rapidly being recognized as viable solutions that provide verifiability, transparency, and stronger data protection without hindering the growth of AI.

The Expanding Issue of AI Trust Deficit

AI is the second most popular category in the crypto space in 2024, attracting over 16% of investor interest. Startups and multinational companies have invested significant resources in developing AI technologies, extending them into people's finances, health, and various aspects of life.

For example, the emerging DeFi x AI (DeFAI) sector has launched over 7,000 projects and reached a market capitalization peak of $7 billion by early 2025, despite a decline following the market crash. DeFAI showcases the transformative potential of AI in making decentralized finance (DeFi) more user-friendly, executing complex multi-step operations through natural language commands, and conducting intricate market research.

However, innovation alone has not addressed the core vulnerabilities of AI: hallucinations, manipulation, and privacy issues.

In November 2024, a user successfully persuaded an AI agent on Base to send $47,000, despite the agent being programmed to never perform such actions. While this scenario was part of a game, it raised genuine concerns: can we trust AI agents to autonomously handle financial operations?

Audits, bug bounties, and Red Team assistance reduce risks but cannot eliminate the dangers of prompt injection, logical flaws, or unauthorized data usage. According to a KPMG report (2023), 61% of people still hesitate to trust AI, with even industry professionals expressing such concerns. A Forrester survey cited by the Harvard Business Review found that 25% of analysts believe trust is the biggest barrier facing AI.

This skepticism remains strong. A survey at the Wall Street Journal CIO Network Summit found that 61% of top IT leaders in the U.S. are still experimenting with AI agents. The rest are either still testing or completely avoiding the use of AI agents, primarily due to concerns over reliability, cybersecurity risks, and data privacy issues.

The healthcare industry feels these risks particularly acutely. Sharing electronic health records (EHR) with large language models (LLMs) to improve outcomes is promising, but the legal and ethical risks are significant without robust privacy protections.

For instance, the healthcare sector suffers adverse effects from data privacy breaches. The problem is exacerbated when hospitals share EHR data to train AI algorithms without protecting patient privacy.

Decentralized, Privacy-Preserving Infrastructure

J.M. Barrie wrote in "Peter Pan": "The whole world is made of trust, faith, and pixie dust." Trust is not just an additional element of AI—it is foundational. Without trust, the projected $15.7 trillion economic dividend from AI by 2030 may never be realized.

At this point, decentralized cryptographic systems have emerged, such as zero-knowledge succinct non-interactive arguments of knowledge (ZK-SNARKs). These technologies offer a new path: allowing users to verify the correctness of AI decisions without revealing personal data or the inner workings of the model.

By applying privacy-preserving cryptographic techniques to machine learning infrastructure, AI can become auditable, trustworthy, and respectful of privacy, especially in fields like finance and healthcare.

ZK-SNARKs rely on advanced cryptographic proof systems that allow one party to prove a proposition is true without revealing the proof process. For AI, this enables models to be verified without disclosing training data, input values, or proprietary logic.

Imagine a decentralized AI loan agent. It does not need to review complete financial records but can check encrypted credit score proofs to make autonomous lending decisions without accessing sensitive data. This protects user privacy while reducing institutional risk.

ZK technology also addresses the black box issue of large language models. By using dynamic proofs, the accuracy of AI outputs can be verified while protecting data integrity and model architecture. This is a win-win for both users and companies—users no longer worry about data misuse, while companies can protect their intellectual property.

Decentralized AI

We are entering a new phase of AI where better models alone are no longer sufficient. Users demand transparency; businesses need resilience; regulators expect accountability.

Decentralized, verifiable cryptographic technologies provide all three.

Technologies like ZK-SNARKs, threshold multi-party computation, and BLS-based verification systems are not just "cryptographic tools"—they are becoming the trusted foundation for AI. Combined with the transparency of blockchain, they create a powerful new tech stack for privacy-preserving, auditable, and reliable AI systems.

Gartner predicts that by 2026, 80% of companies will use AI. The drive to adopt AI is not just hype or resources. It will depend on building AI systems that people and companies can truly trust.

And this begins with decentralization.

Opinion: Felix Xu, Co-founder of ARPA Network and Bella Protocol

This article is for general informational purposes only and should not be considered legal or investment advice. The views expressed in this article are solely those of the author and do not necessarily reflect or represent the views of Cointelegraph.

Related: SEC plans to hold four more cryptocurrency roundtables on trading, custody, tokenization, and DeFi.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。