From the Lecture Hall
to the Real World
IC3 bridges the gap between academic rigor and practical application. We drive the research, hackathons, and global events shaping the frontier of blockchain and AI.
About Us
We are a consortium of academic researchers across 13 universities focused on foundational research in blockchain systems. Our goal is to take concepts from the whiteboard and apply them in the real world.
With expertise spanning applied cryptography, consensus protocols, distributed systems, programming language theory, and game-theoretic mechanism design, our research has played an ongoing, pivotal role in shaping the blockchain industry as we know it.
Featured Research
IC3 publishes research on blockchain, cryptography, decentralized systems, AI and more.
I-HITL: Inalienable Human-in-the-Loop
This paper identifies a critical vulnerability in AI oversight: as AI agents become more capable, they can autonomously circumvent traditional human-in-the-loop (HITL) safeguards — demonstrated by frontier models independently retrieving email OTPs to raise their own spending limits. To address this, the authors introduce Inalienable HITL (I-HITL), a stronger guarantee that human approval remains genuinely required for protected actions even if an AI has compromised the surrounding software environment, with a stricter “strong” variant that blocks even intentional delegation of approval authority to an agent. Using formal communication topology analysis, they prove that a physically trusted component is a necessary condition for real human control, and they validate this with Countersign, a prototype that routes critical approvals through consumer mobile devices as a trusted out-of-band channel — making it practical to retrofit genuine human oversight into existing agentic systems…
RANDAO Manipulation in the Presence of MEV
Ethereum’s randomness beacon (RANDAO) is well-known to be manipulable, and prior work [2] computes the precise fraction of blocks a strategic proposer can propose. The fraction of blocks proposed, however, is only a proxy for participants’ rewards. We propose a generalized reward model capturing many canonical forms of rewards: consensus reward rollover, multiblock MEV, CEX DEX arbitrage, oracle manipulation, and others. We provide a methodology that computes an ε-optimal strategy for any reward scheme in our model (and in particular, any combination of the above rewards). Finally, we apply our methodology to several canonical examples, and establish the sensitivity of RANDAO manipulation to the underlying rewards. We find that if rewards partially roll over, or scale super-linearly with consecutive blocks, the incentive to manipulate RANDAO is amplified. Lastly, we investigate tail-slot slashing, which can be modeled as a reward function, and show that honest equilibria can be recovered…
Holding Agents Accountable with Verifiable Agent Insurance
This paper shifts focus from preventing AI agent misbehavior to recovering from it, proposing Verifiable Agent Insurance (VAI) — a framework where agent providers make financially binding commitments to behavioral policies encoded in smart contracts on a blockchain, backed by monetary escrow. When a violation occurs, an automated verifier reviews execution traces logged through a shared tool-execution path, and confirmed violations trigger escrow payouts to the reporting user. Evaluated across three agent benchmarks, VAI detected violations in 42.2% of violation-containing runs with current agents, rising to 83.0% when agents explicitly emitted structured records of security-relevant actions rather than burying that information in natural language — suggesting that agents and their security policies should be co-designed for auditability from the ground up…
Events & Hackathons
IC3 events connect people across the blockchain industry to share ideas, discuss innovations, and help shape the future of blockchain technology.
IC3 Winter Retreat 2027
January 12-14, 2027
The Winter Retreat will look and feel a little different in 2027. That’s because we’re heading to Cascais, Portugal!
The Science of Blockchain Conference 2026
July 27-29, 2026
SBC focuses on technical innovations in the blockchain ecosystem, and brings together researchers and practioners working in the space.
IC3 Blockchain Camp 2026
June 1-7, 2026
The IC3 Blockchain Camp is one of our biggest events of the year, and the 2026 edition did not disappoint. Princeton University was the main venue for this year’s event.
IC3 Partners
IC3 acknowledges and appreciates a generous gift from the VMware Foundation to advance the science and technology of blockchains.
Get in Touch
Learn more about IC3, partnership, events, and PhD opportunities. Sign up to join our mailing list and stay up to date on all things IC3!