What is it about?
Web 3.0 promises a decentralized internet where users control their own data, but this vision only works if we can trust that data is accurate and hasn't been tampered with. Without proper data integrity—the assurance that information is correct, complete, and unchanged—artificial intelligence systems become unreliable and potentially harmful. This paper explores how ensuring data integrity is not just a technical requirement but fundamental to making Web 3.0 and AI systems trustworthy. We examine current challenges in verifying data across decentralized networks and propose approaches for maintaining truth and accuracy in a world without central authorities. Just as AI without integrity means "artificial" without intelligence, we can't get to Web 3.0 without data integrity.
Featured Image
Photo by GR Stocks on Unsplash
Why is it important?
This paper addresses a critical gap as Web 3.0 technologies rapidly deploy without adequate integrity safeguards. We're the first to comprehensively examine how the absence of data integrity fundamentally undermines both Web 3.0's decentralization promises and AI's decision-making capabilities. Our timing is crucial: as organizations rush to implement blockchain, decentralized AI, and other Web 3.0 technologies, they often overlook that these systems are only as trustworthy as their underlying data. We demonstrate that data integrity isn't just a technical feature but the foundational requirement that determines whether these technologies deliver on their promises or become sophisticated tools for spreading misinformation. This work provides essential guidance for developers, policymakers, and organizations before flawed implementations become entrenched standards.
Perspectives
Writing this piece felt particularly urgent as I watch the tech industry repeat familiar dangerous mistakes from the history of intelligence, with new technologies. Having observed decades of failures, and studied centuries more, I'm struck by how AI evangelists often overlook data integrity or consider trust a solved problem or technical detail. But here's what keeps me up at night: we're building intelligence as automated decisions about people—life and death stuff—on potentially corrupted data in opaque systems with no accountability. "AI without integrity is just A" isn't just a pun; it's meant to be a shrill warning. As Wollstonecraft wrote in 1792, knowledge is power—this shouldn't surprise anyone, yet here we are, watching power structures use deception to masquerade as knowledge. I hope this article helps avoid entrenchment in systems of privilege that leverage deception to deny truths. If we get this wrong, we're not just enabling bad engineering practices—we're undermining the very achievable possibility of a necessary regard for truth in our digital age.
Davi Ottenheimer
London School of Economics and Political Science
Read the Original
This page is a summary of: Web 3.0 Requires Data Integrity, Communications of the ACM, March 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3723438.
You can read the full text:
Contributors
The following have contributed to this page
