Does the default validator list make the XRP Ledger centralized?

The full answer

The XRP Ledger has 150+ validators, according to XRPL.org’s FAQ, but only the ones a server trusts count toward consensus for that server.[2] Servers trust a default list unless their operators change it. That list is where the centralization argument starts, and XRPL.org’s own documentation supports part of it: publishers of these lists “do wield significant power in selecting which validators are widely trusted.”[1]

Who publishes the default lists, and how much do they overlap?

XRPL.org says: “Currently, the default configuration for XRP Ledger servers uses two lists: one published by the XRP Ledger Foundation, and one published by Ripple. Typically, these lists are very similar to one another or even identical.”[1] The FAQ counts “35+ on the default Unique Node List.”[2] In September 2025 the Foundation’s list moved to a new web address and signing key, after stewardship of the ledger passed to a new XRPL Foundation.[3] XRPL.org’s September 18, 2025 migration post says the XRP Ledger Foundation publishes a default UNL “with a list of vetted, independent node operators that exist in a diverse selection of countries, data centers, and so on.”[3]

So two organisations choose the roughly 35 validators that servers listen to by default, and their choices usually match. No dated comparison of the two lists as they stand today is publicly available.

How much power does a list publisher have?

A publisher decides whose votes the servers that follow its list will count. XRPL.org says “Each publisher can define their own criteria for being listed.”[1] The criteria it describes include running “a validator with high uptime for at least a year” and being “a separate and recognizeable entity in the XRP Ledger community.”[1]

That choice reaches upgrades too. Rule changes pass when they have “more than 80% support for two weeks” from trusted validators.[6] Operators of validators “configure their servers to vote on each amendment.”[6] A server that does not upgrade after an amendment passes becomes amendment blocked and can no longer “participate in the consensus process.”[6] Put together, whoever picks the default list picks the electorate for upgrades on every server that keeps the default. Crypto.news reported that XRPL’s official amendment rules state that installing software containing an amendment only gives a server the code needed to understand the proposed rules, and that validators separately choose whether to vote for activation. The upgrade process is set out on how XRP Ledger amendments are approved.

What do defenders of the design point to?

The FAQ says “Ripple does not own or control the XRP Ledger or XRP Ledger network,” and that “Validators can choose to not use the dUNL, or any widely-used UNL for that matter. Anyone can create a new UNL at any time.”[2] XRPL.org also says “Each server operator has full control over which validators are in their UNL.”[1] On the default list itself, the FAQ says Ripple ran “only 1 of the 35 validators” as of July 2023.[2]

The limit on that freedom is overlap. XRPL.org says that “in the worst case scenario, 90% overlap was required to prevent a fork.”[1] An operator who builds a very different list takes on the risk of splitting from everyone else. Whether splits or halts have happened in practice is covered on the XRP Ledger’s outages and incidents.

How has Ripple’s share of the default list changed over time?

Ripple once ran more of the recommended validators. In February 2018 it said: “for every two independent validators that are added in the recommended UNL, one Ripple-operated validator will be removed.”[4] In October 2022 The Crypto Basic reported that “the Ripple operated validators is now at 2” on the Foundation’s list, quoting the Foundation that “our list remains at 35 validators.”[5] The FAQ’s figure for July 2023 is 1 of 35.[2] The primary Foundation posts for 2022 and 2023 could not be opened for this page, and no count after July 2023 was confirmed.

Ripple’s validator count is only one lever. Ripple is also one of the two default list publishers,[1] and the server software’s repository on GitHub says “A team of full-time developers at Ripple & other organizations maintain and continually improve the XRP Ledger’s underlying software.”[8] Who pays for that work is on who funds the XRP Ledger’s software.

Is the consensus protocol itself sound?

This is a separate criticism from who holds the list, and it comes from academic work. In a paper for the OPODIS 2020 conference, Ignacio Amores-Sesar, Christian Cachin and Jovana Mićić described the protocol from its source code and wrote that it “may violate safety and liveness in several simple executions under relatively benign network assumptions.”[9] The paper dates from 2020.

A later chapter published by Springer in 2026, by R. K. Shyamasundar’s chapter, published by Springer in 2026, makes a related argument: that even with the required overlap raised above 90%, safety and liveness are not guaranteed, and that operators need to check how closely every pair of lists matches, rather than only checking each list against one reference.[7] As of September 29, 2026, no public response from Ripple or the XRP Ledger Foundation to either work is known.

How does XRPL’s validator concentration compare with other chains on a common measure?

No like-for-like public measure is known. On the XRP Ledger, the relevant power is being on the trusted list and choosing who is on it, which does not reduce to a share of anything. One way to compare would be to count, for each network, the smallest group of operators able to halt it. No one appears to have published that count.

What does this mean for an XRP holder?

Validator concentration and ownership concentration are different questions. How much XRP sits with a few holders is on who owns the most XRP. Whether Ripple can freeze or take anyone’s XRP is on whether Ripple controls XRP.