Ask Finn← Discover
WORTH KNOWING

Cornell Wins $10 Million to Fight AI's Growing Threat to Online Trust

By Hayden Walsh · Saturday, November 7, 2026
Finn's Take· TL;DR
  • Cornell secured $10M to develop auditable AI infrastructure that lets online communities control how AI participates in discussions while maintaining transparency.
  • The team is building three-part trust system: AI assistants for clarity, moderators for group discussion, and independent auditors using cryptographic verification for accountability.
  • Tools like Assistant Toolkit and Auditing Benchmark enable communities to test AI behavior independently rather than trusting tech companies' claims about their systems.
See this from any side — with sources:
Left takeNeutralRight take

A Crisis of Trust in the Digital Age

"Right now, our shared public sphere is entering a crisis of trust," as one researcher put it. "People no longer believe in what they read, they don't know who they are talking with, and perhaps more , they don't trust the AI systems that are mediating the exchanges." That diagnosis now comes with a serious funding commitment to fix it.

A multi-year, $10 million "moonshot" grant from Laude Institute will help Cornell researchers realize the potential of AI, while giving online communities control over its role. The team is one of four that received moonshot awards from Laude Institute to leverage AI to solve humanity's biggest problems, out of eight finalists. The project is formally titled "Rebuilding Trust in Civic Discourse: An Infrastructure for Auditable AI."

What the Research Team Plans to Build

The interdisciplinary team of Cornell faculty and students, with expertise in AI, conversational behavior, computational systems and human-centered design, will develop tools and protocols to rebuild the foundations of trust online, where humans are no longer the only participants. They are working directly with online communities, including those on Wikipedia, Reddit, and Discord, to understand their communication needs and co-design tools that reflect community values and governance practices.

The project proposes a three-part AI trust infrastructure for online civic conversation — assistants that help individuals articulate views clearly, moderators that support group discussion, and an independent auditor that uses cryptographic verification to ensure the whole system stays neutral and inspectable from the outside. Among the specific tools already developed is an Assistant Toolkit. The Assistant Toolkit allows community members to build, configure and test AI tools that participate in group discussions or privately support individual users, reflecting their community's needs and rules. ConvoArena and Simulation Factory support studies with human participants and tests using simulated conversations to investigate how AI interventions affect discussions. The Auditing Benchmark provides a shared testing ground for developing and comparing independent auditors that check whether AI assistants follow their stated specifications.

The researchers aim to make interventions transparent and accountable rather than ask communities to accept a system's promised behavior on trust. That distinction matters enormously. In an era where AI systems are increasingly embedded in everyday platforms, the ability to independently verify what those systems are actually doing — rather than simply taking a tech company's word for it — represents a fundamental shift in how AI accountability could work.

The People and the Money Behind It

"We are at a turning point in the history of human communication. AI is becoming an active participant in the conversations through which we form opinions, resolve disagreements and make collective decisions," said Cristian Danescu-Niculescu-Mizil, professor of information science in the Cornell Ann S. Bowers College of Computing and Information Science and lead principal investigator on the project. The team also includes Natalie Bazarova, Jon Kleinberg, Robert Kleinberg, Mor Naaman, and David Rand.

Moonshots is a flagship initiative of the Laude Institute, which was founded by Andy Konwinski — a computer scientist who co-founded Databricks, a cloud-based platform for data analytics, and Perplexity, an AI search engine — to accelerate the transfer of frontier computer science research into real-world impact. "We are at a moment in human history in which we have more power as a species than we've ever had," said Konwinski at the award ceremony.

Why This Could Matter Beyond Academia

"Recent advances in AI pose a serious threat to civic discourse, enabling deepfakes, impersonation and manipulation," said team member David Rand. "But AI also has the potential to be a real positive force. In our experiments, conversations with AI can help correct misconceptions and mediate difficult conversations. Our project aims to unlock that positive potential in a way that communities can actually verify and trust."

The goal is not for everyone to agree or trust each other, but to ensure the environment itself supports meaningful and trustworthy interactions, even in the presence of AI. That's a more modest — and arguably more realistic — ambition than trying to eliminate online conflict altogether. If the Cornell team succeeds, the tools they build could eventually serve as a template for how any online platform, from neighborhood forums to national news comment sections, manages AI's growing role in shaping what we say, read, and believe.

Have a question about this story?
Ask Finn — answers grounded in this article, from any viewpoint.