How was the XRP Ledger's native cross-chain bridge (XLS-38) designed to link two ledgers, and is it live as of October 13, 2026?

The full answer

What did XLS-38 propose?

XLS-38 set out a protocol for bridges built into the XRP Ledger, aimed at custom sidechains. RippleX Developers summarized it on August 27, 2026: instead of relying on third-party bridge operators, a decentralized network of witness servers would observe and attest to transactions across chains, so assets could move between XRPL mainnet and any connected sidechain [1].

The feature lives in the XChainBridge amendment. XRPL.org’s documentation on cross-chain bridges says they move XRP and tokens between the XRP Ledger and other blockchains, and that they require that amendment [2].

RippleX described the original aim as giving developers building blocks for private chains, permissioned networks and test environments for features not yet on mainnet [1]. It also said XLS-38 was meant to connect the XRPL EVM Sidechain to mainnet, with XRP as the sidechain’s gas token [1].

Why did RippleX recommend withdrawing it?

The post, published by RippleX Developers on August 27, 2026, gave three reasons [1]. The EVM Sidechain, the main use case behind XLS-38, is served by Axelar instead. RippleX wrote that no private sidechain demand had materialized for projects that would use XLS-38 specifically [1]. And keeping the code in xrpld carries a maintenance cost; RippleX estimated that withdrawal would let it remove more than 10,000 lines of code [1].

The security argument sits underneath those reasons. RippleX wrote that the witness model puts a great deal of trust in a set of operators: a larger, more decentralized set improves security but makes coordination and governance harder [1]. For a bridge securing a public chain’s native gas token, Ripple concluded it lacked the operational expertise to run this well at scale. It said Axelar has 75+ validators, a threshold signature scheme and experience across 55+ blockchains [1].

RippleX also said the recommendation covers fixXChainRewardRounding, a related amendment [1]. The post does not describe how witness servers would have been paid.

Is the XChainBridge amendment live on mainnet?

RippleX’s own post treats it as switched off. Ripple said that in June 2024 it committed to leaving XLS-38 open for a vote while its UNL validator kept voting No, and gave the community 12 to 15 months to show interest [1]. dev.to stated on August 27, 2026: “We didn’t find the level of adoption or developer interest that would justify keeping the amendment moving toward activation.”

Amendments take effect only through validator voting, a process explained in how XRP Ledger upgrades get approved.

How would the withdrawal work?

RippleX laid out three steps on August 27, 2026 [1]. Ripple said in its August 27, 2026 recommendation that its intended first step is to open a pull request to the xrpld repository that sets VoteBehavior::Obsolete for the XChainBridge feature, and that once all active validators consider the amendment Obsolete, it can be removed from the codebase entirely. Second, the amendment would lose support as validators upgraded. Third, once all active validators treat it as Obsolete, a later update would remove the XChainBridge and fixXChainRewardRounding code from xrpld [1].

The post said a comment period would stay open before moving ahead, with feedback invited on GitHub and the XRPL developer Discord [1]. How much weight Ripple’s own software choices carry in that process connects to the debate over whether the default validator list makes the ledger centralized.

What changes for private networks and the EVM sidechain?

RippleX wrote that builders on the XRPL EVM Sidechain see no change, because the Axelar bridge remains their route to and from mainnet [1]. The sidechain itself is covered on the page about XRP Ledger smart contracts and the EVM sidechain.

For private or permissioned chains, RippleX said no single mechanism fits every case. It named Axelar and Wormhole as one class, alongside ZK-based approaches, L2-style designs and research such as consensus-within-consensus, and said none is a drop-in replacement for XLS-38 [1].

What is the strongest case against witness-based bridges, and for keeping XLS-38?

Bridges have been a frequent target for theft. Chainalysis estimated on August 2, 2022 that $2 billion had been stolen across 13 cross-chain bridge hacks, and that bridge attacks made up 69% of all funds stolen in 2022 to that date [3]. RippleX’s point that witness-based security depends on trust in a set of operators is the specific worry for XLS-38 [1].

The case for keeping it comes from RippleX’s own wording. The company said it would revisit the recommendation if developers present concrete plans, and that the design work “wasn’t wasted” [1]. A separate question is whether new ledger features translate into value for XRP at all, weighed on the page about XRP’s valuation versus ledger activity.