Safeguarding DeFi: Managing Risk in a Decentralized Economy

In the REDeFiNE TOMORROW 2025 event in a session"Safeguarding DeFi: Managing Risk in a Decentralized Economy," moderated by Melody He, Managing Partner and Co-Founder of The Spartan Group, Omer Goldberg, CEO and Founder of Chaos Labs, presented a vision and technology that is becoming a critical infrastructure for the DeFi world by transforming complex risk management into an automated, AI-driven system.

The Original Problem and the Solution The Financial Risk Platform

In the rapidly growing world of DeFi, a significant challenge is managing ever-changing risk. Chaos Labs has stepped in to solve this problem with its Financial Risk Platform, an AI-driven risk management platform that serves as the brains behind numerous leading protocols such as Aave, Jupiter, GMX, dYdX, Ether.fi, Renzo, and LayerZero.

The platform calculates and streams AI-generated risk parameters on-chain in real-time across over 20-30 blockchain networks. For example, if you use Aave, critical parameters like which assets can be listed, Loan-to-Value ratios, liquidation thresholds, or interest rates are constantly determined and updated by Chaos Labs' technology.

The Evolution to Chaos Oracles When Data Needs Context

In the past, modifying parameters in Smart Contracts was a manual process that was slow, error-prone, and made protocols inflexible. To solve this, Chaos Labs created Chaos Oracles, which consist of three main components:

1. Price Oracles: Report price data.
2. Risk Oracles: Stream financial risk parameters (Chaos Labs is the only provider in the market).
3. Proof of Reserves: Increases system transparency.

The key differentiator of Chaos Price Oracles is that they don't just report raw price data; they also consider "depth of liquidity" and "price impact," using a "business logic" from an AI-driven Risk Engine to interpret the data before delivery, which is completely different from typical oracles.

The importance of this context can be seen in past events, such as when a Chainlink feed on Avalanche caused a momentary error, resulting in over $500,000 in losses for Euler Protocol users, or the Mango Markets attack on Solana, where an attacker manipulated the price of a low-liquidity token to borrow vast sums of assets. These incidents demonstrate that a lack of in-depth risk assessment can cause hundreds of millions of dollars in damages and represent a critical vulnerability arising from the fact that many DeFi builders may lack financial experience.

The Next Big Step Chaos AI

At the heart of Chaos Labs is the development of Chaos AI, born from the belief that calculating real-time financial parameters requires statistical models driven by Machine Learning. It started with an Agent-based Monte Carlo Engine and was later enhanced after the launch of GPT-3, which allowed them to extract deep insights from "unstructured data" like text and news.

A key strength is their "Proprietary Data Lake," which collects massive amounts of data from blockchains, CEXs, and traditional financial markets, giving Chaos AI a "data advantage" that general models lack. Furthermore, the architecture of Chaos AI is not a single model but an "AI System" composed of 15-16 collaborating sub-models, each with a specific expertise, such as capturing intent, writing responses, or performing calculations.

The incredible interest in Chaos AI is confirmed by the massive response:

• Over 1 million people are on the waitlist.
• An initial rollout has begun with the first 100 users, with a broader release planned for the summer.
• It is estimated that if released to everyone on the waitlist, the inference costs would be as high as $500 million to $1 billion annually.

Philosophy and Vision: Connecting the Financial World and the Path to Decentralization

Omer Goldberg believes that On-chain Finance and Web2/Traditional Finance are practically no different if built with the right data structure. AI must be able to understand and reason about both asset types as "first-class citizens," which is the core concept of AFI (Artificial Financial Intelligence). This perspective is crucial, considering that 10-15% of the global crypto market cap is now in the form of tokenized assets; therefore, a good AI must understand both worlds.

In terms of decentralization, Chaos Labs takes a pragmatic approach, prioritizing Product-Market Fit first, as decentralizing too quickly can lead to inefficiency. Omer points out that decentralization is a "spectrum," not a binary on-off switch, and noted with an interesting perspective: "It's all fun and games until it really matters," reflecting the necessity of building a robust and practical product first.

With a clear vision, claims of being the fastest-growing Oracle in the market, and new partnerships to be announced in the coming weeks, Chaos Labs is positioning itself as a leader in the risk infrastructure that will power the next wave of DeFi innovation safely and sustainably.

Summary

Chaos Labs is driving a revolution in DeFi with its AI-powered risk management platform, enhancing security for leading protocols. They developed Chaos Oracles that analyze price, liquidity, and transaction impact, incorporating "business logic" to prevent damage from attacks. For them, risk assessment is the highest priority. Chaos AI, built on a Proprietary Data Lake, enables the AI to think and reason about financial data with precision. Although building AI in finance is challenging, Chaos Labs utilizes a system that integrates multiple models. They prioritize Product-Market Fit, believing that an excellent product will naturally lead to the decentralization of AI and foster a stronger DeFi future.