
New protocol rules for transaction handling and AMMs are now live, requiring infrastructure operators to update their software to maintain network compatibility.
AI-generated summary
The XRPL amendment system allows for network upgrades to be voted on by validators. Once an amendment reaches a majority, it becomes a mandatory rule for all network participants.
The XRP Ledger activated a package of fixes for transaction handling, AMMs, and newer protocol features, putting outdated servers at risk.
The fixCleanup3_3_0 amendment went live Sept. 11 at 11:29 UTC in validated ledger 106,911,489, according to XRPSCAN. Ripple’s public XRP Ledger endpoint showed the amendment enabled and supported on Sept. 12.
The update bundles maintenance corrections involving Checks, automated market makers, pseudo-accounts, permissioned trading, Single Asset Vaults, and the Lending Protocol as part of the network’s live rules.
Several changes target transaction paths that could previously fail in edge cases. CheckCash and CheckCancel now reject an all-zero CheckID during preflight with temMALFORMED, while a narrow divide-by-zero condition involving AMMWithdraw returns tecAMM_FAILED instead of ending in tefEXCEPTION.
Other fixes adjust freeze handling for transfers involving pseudo-accounts and tighten safeguards around AMMs, permissioned markets and cleanup processes, according to the XRPL amendment registry.
The code supporting those changes has been available since xrpld 3.3.0 was released Aug. 6. Their activation this week marks the point when they shifted from optional software support into rules that servers must understand to remain fully compatible with the ledger.
The fixes now carry consequences for outdated servers
XRPL’s amendment system allows new code to be distributed before validators make it part of the network’s consensus rules. Once an amendment secures the required majority for the prescribed period and activates, servers that do not support it can become amendment blocked.
That creates the immediate operational issue for exchanges, custodians, payment companies and other businesses running their own XRP Ledger infrastructure.
XRPL documentation says an amendment-blocked server cannot determine whether ledgers are valid, submit or process transactions, or participate in consensus. Restoring those functions requires software that supports the active amendment, with the project recommending operators use the latest appropriate xrpld release.
XRPSCAN reported 31 of 35 validations supporting fixCleanup3_3_0 against a threshold of 28 at 01:45 UTC on Sept. 12, after the amendment completed its two-week majority period.
The activation does not mean every feature included alongside the fixes in version 3.3.0 is now live.
Separate amendments covering LendingProtocol, SingleAssetVault, BatchV1_1, ConfidentialTransfer, DynamicMPT, PermissionDelegationV1_1 and Sponsor remained disabled on Sept. 12. Ripple’s public server reported support for those amendments, meaning the software can recognize them even though validators have not yet made them active network rules.
For users relying on maintained wallets or exchanges, the compatibility work largely sits with service providers. Firms operating their own XRPL servers now face the more immediate requirement of keeping those systems current as validator-approved amendments move from available code into the ledger’s active rule set.

OpenAI's new ChatGPT Images 2.5 model is compared against Google's Nano Banana 2. While OpenAI resolved previous oversharpening issues, the models remain closely matched in performance, with each showing specific strengths in text rendering, spatial awareness, and factual accuracy.

Greenberg Traurig reported an unauthorized actor accessed and posted limited documents on the dark web, exposing client Social Security information. The breach is part of a growing trend targeting law firms, with BakerHostetler handling nearly 60 cybersecurity incidents in 2025. Other firms including Taft Stettinius, Herbert Smith Freehills, WilmerHale, Goodwin Procter, and Quinn Emanuel disclosed breaches. Crypto companies Coinbase, Ledger, and SafePal also reported customer data exposures.

OpenAI has consulted members of Congress on whether rival AI companies could legally agree to slow development, following internal calls for shared safety standards amid intense commercial and geopolitical competition. The outreach comes as researchers warn of pressure to cut corners on safety, with concerns that unilateral slowdowns put companies at a competitive disadvantage without industry-wide coordination.

AI agents have reduced the computational cost of a potential quantum attack on Bitcoin's secp256k1 signature scheme by 86% in two months. Researchers from major crypto firms published these findings to accelerate defensive planning against future quantum threats.

Anthropic reported that Russian- and Chinese-speaking operators, as well as an independent consultant linked to Mali's intelligence, used its Claude AI model to automate cyberattacks, vulnerability research, and a population-scale surveillance system.

Blockstream refused a ransom demand from hackers holding about 598 Bitcoin on the Liquid Network, labeling the theft a crime rather than responsible disclosure and vowing to work with law enforcement.