Ethlabs, a nonprofit founded by former Ethereum Foundation researchers earlier this year, proposed advancing Ethereum’s post-quantum upgrade to 2027. The group argues that AI is accelerating quantum and cryptographic risks, making earlier action necessary to secure the network as a credible decentralized root of trust. They named the proposal "pqai27" in a post on the Ethereum Magicians forum.
The AI Acceleration Argument
The proposal cites the U.S. government moving its quantum risk goal from 2035 to 2030 and Google aiming for post-quantum status by 2029. The Ethereum Foundation previously set December 2029 for full readiness, but Ethlabs contends that AI now solves hundreds of open mathematics problems and finds roughly 1,000 vulnerabilities in Linux. "Being post-quantum is not enough; we intend on being post-AI too," Ethlabs wrote. This definition includes cryptography with minimal assumptions and clients hardened against frontier models using formal verification.
I find the urgency convincing because the threat landscape has shifted from theoretical to immediate. Ethlabs notes that it is now much harder to reason when any specific threat will arrive. They state that even at current levels, AI accelerates all these risks. The group warns that security is itself an open maths problem, so waiting until 2029 leaves a dangerous gap.
Technical Path and Trade-offs
Ethlabs distinguishes two levels of readiness for 2027: "pqai-ready," where consensus is post-quantum and users have a post-quantum route for accounts while the network runs at scaled capacity, and "pqai-scaled," where the network operates at full capacity with post-quantum transactions and data availability. To meet this target, upcoming hard forks like Hegota must be reassessed. Ethlabs will withdraw quick slots, a proposal to speed up block production, from Hegota to dedicate resources to pqai27.
This trade-off is painful. Ethlabs acknowledges they pushed for quick slots because community surveys showed clear demand. However, they argue that pivoting later, deep into shipping the currently planned fork, would be far more costly than acting now. They claim Ethereum already has a clear technical path built around hash-function-based cryptography and embedded formal verification practices.
"With attacks on software taking off, security and uptime will be more in demand than ever," the proposal said. It asserts that resilience against AI-led uncertainty will be Ethereum’s next killer feature. I agree that committing now is cheaper than retrofitting later. The decision to drop quick slots shows a clear priority shift toward long-term survival over short-term speed.
