AI agents are changing the game in decentralized finance, or DeFi. They automate trading and liquidity processes, making things faster and more efficient. But there are some important questions about data accuracy and safety.
These AI agents are now a key part of DeFi. As the trend of automating financial systems grows, their safety is under the spotlight. How safe are these systems? That’s what many are asking.
To work effectively, AI agents need precise data. They use this data to automate liquidity provisioning, execute trades, and manage portfolios. As more people adopt these technologies, concerns about their reliability and safety become increasingly critical.
Mike Cahill, the CEO of Douro Labs and a contributor to the Pyth Network, explained that AI agents depend on “real-time, high-fidelity data to make split-second decisions.” He pointed out that errors or manipulation in the data can lead to serious consequences. To address these risks, he emphasizes the need for “ultra-low-latency, first-party price updates.”
AI agents rely on accurate and timely data to make informed decisions in fast-moving markets. Ideally, this data comes directly from first-party providers, like market makers or exchanges. By aggregating data from multiple sources, these systems reduce the chances of manipulation or inaccuracies. Updates can happen in as little as one millisecond, allowing agents to respond instantly to market changes.
Cahill noted, “Pyth ensures agents operate on the most accurate market data available—eliminating risks tied to stale or manipulated information.” He added, “AI agents thrive on speed, precision, and automation.”
However, one of the biggest challenges for AI agents is ensuring they operate safely in volatile market conditions. Decentralized systems aim to address this concern. For instance, Oracle Integrity Staking (OIS) requires data publishers to stake capital. This aligns their financial interests with the accuracy of their data contributions. If they provide faulty or manipulated data, they risk losing their stakes.
Cahill explained that Pyth’s OIS creates an “economic security layer.” This layer works alongside first-party price sourcing and weighted aggregation to develop resilient, high-frequency pricing that reflects true market conditions. He also mentioned that “AI agents can integrate programmable safeguards, like confidence intervals and predefined slippage thresholds, preventing them from executing trades in volatile or unreliable conditions.”
Looking ahead, Cahill envisions a future where “fully autonomous financial systems operate more efficiently than any human-run market ever could.” He expects to see artificial general intelligence (AGI) agents within the next one to three years.
“Real-time data gives AI agents the ability to unlock a new era of high-frequency, algorithmic trading in DeFi,” Cahill said. “This is where institutional DeFi surpasses traditional finance, creating a market that’s not only decentralized but also faster, more efficient, and truly autonomous.”
The growing interest in AI agent technology is evident in recent developments in the crypto industry. For example, Fetch.ai has launched a $10 million accelerator for AI agent startups. Additionally, the developer behind ai16z has published a white paper outlining a vision for Web3-native AI agents.