An AI security tool has identified a critical flaw in a proposed XRP Ledger (XRPL) upgrade, potentially preventing significant fund losses. The vulnerability, discovered by security researcher Pranamya Keshkamat using the AI auditing tool Apex, was found in the network's proposed Batch Amendment before it went live. This prompted developers and validators to halt the amendment's activation and implement emergency safeguards. The flaw, identified on February 19, 2026, involved a logic error in the validation of batch transaction signers, which could have allowed attackers to move funds without private keys. Ripple's engineering team quickly released a fix, urging validators to vote against the amendment and deploying rippled 3.1.1 to disable the vulnerable features. XRPL Labs has since integrated AI-assisted audits into its review process to prevent similar issues in the future.