Bitcoin (BTC) Dev Paul Sztorc Claims 80% of Lightning Payments Are Custodial

Sztorc’s 80% Custodial Claim

Bitcoin (BTC) developer Paul Sztorc has reopened a long-running dispute over the Lightning Network with a blunt charge: the community has been lying to itself about how the payment layer actually works. In his October 4 post on X, Sztorc wrote that “most people don’t realize how bad the Lightning Network is,” blaming a “vast, years long conspiracy” to misrepresent the protocol’s user experience. His sharpest line compares the promotion of custodial Lightning services, which he claims account for more than 80% of payments, to a “campaign of fraud on par with Elizabeth Holmes’ Theranos.” The claim is hard to check from the outside. Lightning routes BTC through payment channels that settle off-chain rather than on Bitcoin’s proof of work base chain, and payments on the network are private by design, so no public dataset can establish what share runs through custodial intermediaries. He has not backed the figure with public data. The fight matters because the Lightning Network is the flagship of an effort, long championed by parts of the Bitcoin maximalism movement, to make Bitcoin (BTC) spendable for everyday purchases rather than a passive store of value. Even insiders concede friction: routing can fail for larger payments that need several hops through different nodes, and running a node with managed channels stays too technical for most users, which is why custodial services exist at all. Alex Bosworth, the network’s head of liquidity, wrote in a separate X post this month that “the painful truth is that in a setup with privacy and p2p etc it’s extremely slow compared to tap to pay.” While the Bitcoin price keeps most market attention fixed on macro flows, this dispute is about whether the network’s own payment layer does what its builders promised. The argument resurfaces a question developers have wrestled with for years: has the protocol made BTC easier to spend, or has the community’s investment in the layer blinded it to the friction users actually feel?

@Truthcoin · X post

His October 4 post on X.

View on X

Amboss Puts $100 Payment Success at 99.66%

Industry operators have answered with numbers of their own. Jesse Shrader, co-founder and chief executive of Lightning infrastructure provider Amboss, acknowledged that liquidity requirements are challenging to understand at first but called them trivial to solve with current tools or automation services. “For everyday payments of around $100, most users can achieve 99% reliability in the core network,” he said, and Amboss’ simulations put the success rate for $100 payments at 99.66%. He rejected Sztorc’s custodial claim outright: payment statistics at network level cannot be produced unless a party observes the transactions themselves, so in his view nobody can prove an 80% custodial share from public data alone. Danny Stagg, marketing lead at payments infrastructure provider Breez, said the criticism described an earlier era. In his view the network has improved substantially over the past two years, with newer infrastructure abstracting away channel management, and non-custodial user experience is now comparable with custodial products. Breez chief executive Roy Sheinfeld has separately argued that AI coding tools are making it easier to build Bitcoin-enabled applications, with inquiries from prospective partners rising. The measurable record is mixed. Financial services firm River’s 2026 Bitcoin adoption report estimated that Lightning processed $1.17 billion in transaction volume across 5.22 million transactions in November 2025, a count still below the network’s August 2023 peak of 6.6 million transactions. Payment processor CoinGate’s first-half 2026 data showed 782,403 paid crypto orders, up 21.4% year over year, with Lightning accounting for 9.6% of the Bitcoin (BTC) payments it processed. As far back as 2023, River estimated one self-custodial user for every eight custodial users. The wider settlement picture has shifted too: stablecoin volume on other chains, where yield-bearing stablecoin designs reward holders, runs into trillions, and routes that move BTC off-chain as Wrapped Bitcoin, alongside emerging Bitcoin DeFi applications, compete for the same payment flows Lightning was built to capture.

Our reading: neither headline number settles the argument, because the network’s privacy design hides exactly the data both sides need. Sztorc’s 80% custodial share has no public baseline it can be measured against, so its size cannot be verified, while Amboss’ 99.66% success rate is a simulation rather than a network-wide record. What the public record does show is directional: one processor attributes 9.6% of its Bitcoin payments to Lightning, and transaction counts sit below the 6.6 million peak of August 2023 even as the value flowing through the network rose to $1.17 billion in a single month. Until someone produces a measurable base for the custodial share, the fight stays a claim about incentives, not a finding about usage.