Raydium Hit by $1.34M Exploit — Legacy Code at Fault

John NadaBy John Nada·Jun 10, 2026·3 min read
Raydium Hit by $1.34M Exploit — Legacy Code at Fault

Raydium's legacy AMM pools exploited for $1.34M. No current users affected. A lesson in vigilance.

In a significant incident highlighting the persistent vulnerabilities within decentralized finance (DeFi) platforms, Solana-based decentralized exchange Raydium was targeted in an exploit that led to a loss of more than $1.34 million. This exploit, which occurred on a Wednesday, specifically impacted five deprecated liquidity pools from an older version of Raydium's automated market maker (AMM) program. Despite reassurances from the pseudonymous Raydium contributor 0xInfra on X that 'no current users of Raydium are affected by this exploit or would have been able to interact with these pools through the UI since their deprecation,' the incident underscores the critical need for ongoing vigilance in DeFi security.

The exploit was orchestrated by an attacker who exploited vulnerabilities in legacy code, bypassing validation logic to mint new liquidity provider tokens. This allowed them to make off with nearly $900,000 in USDC, approximately $357,000 in Solana (SOL), and $86,000 in Raydium’s native token, RAY. The impacted liquidity pools had been phased out back in 2021, yet they remained as potential attack vectors due to lingering legacy components.

Raydium has assured its users and stakeholders that the current systems in place are impervious to such vulnerabilities, suggesting that the oversight was not a result of a compromise at the authority level, but rather a failure in fully decommissioning obsolete systems. This incident serves as a stark reminder of the importance of not just updating systems, but ensuring that outdated elements are completely eradicated to prevent exploitation.

The broader DeFi landscape has seen a series of similar incidents in recent months, highlighting a trend that is increasingly concerning for stakeholders. In April, for instance, KelpDAO and Solana-based Drift Protocol both suffered exploits that resulted in losses nearing $300 million each. These incidents have sparked discussions about the challenges facing DeFi security, especially as the sector continues to grow and attract more attention from both legitimate users and malicious actors.

The use of artificial intelligence (AI) in vulnerability discovery is adding a new dimension to these challenges. Although there is no specific evidence that AI played a role in the Raydium exploit, AI is transforming how threats are identified and addressed. Analysts highlighted that AI is automating tasks traditionally performed by skilled auditors, potentially accelerating the discovery of vulnerabilities. This development is exemplified by the efforts of companies like Anthropic, which recently released an upgraded version of its cybersecurity-focused AI tool, Mythos. This tool is touted for its 'unprecedented cybersecurity capabilities,' further illustrating the rapidly evolving landscape of cybersecurity in the DeFi space.

In response to the exploit, Raydium has indicated that the financial impact will be cushioned by the firm's treasury, which will cover the losses. Despite this, the market has reacted to the incident, with Raydium's native token RAY experiencing a 2% drop in the last 24 hours. This decline is part of a broader market trend, with the token having fallen around 13% over the past week, bringing it to 96.6% below its all-time high of $16.83.

The incident with Raydium illustrates the broader challenges facing DeFi platforms as they navigate the complexities of maintaining security while continuing to innovate and expand. The lesson is a clear one for DeFi platforms everywhere: vigilance must extend beyond the shiny, new systems to those relics of the past that still linger on, posing quiet threats in the background. Addressing these legacy vulnerabilities is crucial to safeguarding against potential exploits and ensuring the continued growth and stability of the DeFi ecosystem.

Scroll to continue