The Ethereum Foundation is quietly retiring the Poseidon hash function, a cryptographic primitive long favored for zero-knowledge proofs, in favor of the more conventional SHA and BLAKE3 families. The change, confirmed this week, signals a deliberate pivot toward broader compatibility and post-quantum preparedness, even if it means sacrificing some of the efficiency gains Poseidon offered in zk-rollups and other privacy-focused applications.
Why the switch
Poseidon was designed to be zk-friendly, meaning it requires fewer constraints when implemented inside a circuit. That made it a natural fit for projects building on Ethereum's scaling and privacy layers. But the Foundation's move away from it suggests a different set of priorities. SHA-256 and BLAKE3 are already widely deployed across the ecosystem, from consensus protocols to general-purpose cryptography. They're battle-tested, well-audited, and supported by a much larger body of tooling.
The post-quantum angle is harder to ignore. Neither SHA nor BLAKE3 is quantum-safe in the strictest sense, but they're considered more robust against certain attack vectors than Poseidon, which was built with a different threat model in mind. The Foundation hasn't said it's going fully quantum-resistant tomorrow, but the direction is clear: standardize on hashes that are easier to upgrade and harder to break in unexpected ways.
What changes for developers
For teams already using Poseidon in production, this isn't a trivial swap. The hash function sits deep inside circuit logic, and changing it means re-auditing proofs, updating prover code, and potentially rethinking how data is committed. The Foundation's guidance is that new projects should default to SHA or BLAKE3, while existing deployments can migrate gradually. There's no hard deadline attached to the transition, but the message is unambiguous.
The practical upside is that developers get access to a much richer ecosystem of libraries and hardware acceleration. BLAKE3, in particular, is fast on commodity CPUs and has been adopted by a range of file and data integrity tools. SHA-256 remains the baseline for most blockchain work. For teams that don't need the extreme constraint optimization of Poseidon, the trade-off is easy to justify.
Compatibility first
The Foundation's emphasis on compatibility reads as a response to the fragmentation that has crept into Ethereum's cryptographic stack. Different rollups and privacy protocols have been using different hash functions, which complicates cross-chain interoperability and makes it harder for auditors to build reusable tooling. Moving to a common set of hashes is a step toward unifying the ecosystem, even if it means giving up some performance in niche use cases.
It's also a hedge. Post-quantum research is moving fast, and the Foundation wants to avoid being locked into a hash that might not hold up under future cryptanalytic scrutiny. SHA and BLAKE3 have longer track records and more eyes on them. That's a conservative choice, but in cryptography, conservative is often the right call.
The transition won't happen overnight. Existing circuits using Poseidon will keep running, and the Foundation hasn't announced a forced migration. But the direction is set. New specifications and reference implementations will lean on SHA and BLAKE3, and the message to the developer community is to plan accordingly. The next concrete step is likely an updated set of Ethereum Improvement Proposals that formally deprecate Poseidon in favor of the new standards.
