Sablier Shuts Down Development as AI Tools Erode Protocol Moat

Published by James Harris on July 14, 2026July 14, 2026

Sablier Labs, the team behind one of Ethereum’s earliest token-streaming protocols, is shutting down active development. CEO Paul Razvan Berg announced on July 13 that the company is entering maintenance mode, citing a brutal first quarter and no clear path to building a venture-scale business. The protocol itself will keep running until at least June 2028, and the smart contracts have been open-sourced immediately, but no new products are coming and no new chains will be supported.

The proximate cause is straightforward: revenue and usage dropped sharply in Q1 2026 even as the team shipped more features than in any prior quarter. Berg identified two culprits. Token launches, which generate demand for vesting and streaming infrastructure, were delayed as the broader market weakened. And AI-assisted development tools made it easier for teams to replicate what Sablier built, eroding the moat that justified a standalone business.

The Thesis That Didn’t Scale Fast Enough

Sablier was first to introduce on-chain money streaming as a concept, formalizing it through ERC-1620 back in 2019. The underlying bet was that DAOs would proliferate, payroll would move on-chain, and token-denominated compensation would become routine enough to sustain a protocol business. That bet was not entirely wrong. DAOs did grow. Token vesting did become standard practice. But the volumes never reached the threshold Sablier needed, and the use cases that actually generated crypto’s growth in this cycle, speculative trading, prediction markets, decentralized lending, had no particular need for streaming infrastructure.

This is a pattern worth recognizing. Several infrastructure teams that built during 2020-2022 on the assumption that DeFi primitives would compound into mainstream financial rails have quietly wound down or pivoted over the past eighteen months. The demand materialized, just not in the direction anyone mapped. Sablier’s situation is less a failure of execution than a mismatch between the infrastructure layer they built and where on-chain activity actually concentrated.

An Early License Release as the Real Legacy

The more consequential decision may be the licensing change. Sablier’s core EVM contracts were originally under Business Source License 1.1, with a scheduled conversion to GPL in July 2029. Berg accelerated that to July 13, 2026, meaning developers can fork, modify, and redeploy the contracts right now rather than waiting three more years. Bankless called this the project’s “parting gift” to the community, which is accurate in a narrow sense, but it also reflects something more structural: when a protocol cannot sustain itself as a business, open-sourcing the IP is often the only way to preserve its technical contribution.

For teams currently using Sablier for vesting schedules or airdrop distributions, the announcement is clear that nothing breaks immediately. Smart contracts are permissionless and non-custodial, so existing streams continue to execute on-chain regardless of what the company does. The interface will stop supporting new vesting streams and airdrops after June 2028, and open-ended payment streams are already disabled as of the announcement date.

What the Shutdown Signals for Infrastructure-Layer Projects

Sablier’s exit is a useful data point for how this cycle is sorting winners from casualties. Projects that built financial primitives expecting DAO adoption to drive volume are facing a reckoning: the DAO ecosystem exists, but it has not scaled to the economic activity levels that would make payroll-style streaming a high-volume business. Meanwhile, the teams building on top of speculation, leverage, and liquidity provision have captured the actual demand.

Berg’s framing, that June 2028 marks the end of company-funded maintenance rather than the death of the protocol, is worth taking at face value. The contracts are live, the code is now open, and the ERC-1620 standard remains part of Ethereum’s history. Whether a community-run fork emerges depends entirely on whether any team finds the streaming primitive useful enough to maintain without a business model attached. Given where on-chain activity is concentrated right now, that is an open question.