Phala Leaves Polkadot For Ethereum L2

Key Takeaways

• Phala is officially migrating from Polkadot to Ethereum Layer 2.

• Tokenholders will receive a 1:1 ERC-20 version of PHA before Nov. 20, 2025.

• The move is aimed at better scalability, developer access, and enterprise-grade AI compute.

• Unlike other Polkadot projects, Phala is fully leaving the ecosystem, not going multichain.

• Ethereum L2 aligns with Phala’s strategy for confidential GPU-based computing and AI agent infrastructure.

A Community-Backed Move To Ethereum L2

On October 9th, Phala’s community passed a major proposal confirming its full migration to Ethereum L2. The transition, already in motion, will be completed before November 20th, 2025.

Source: Phala Network

Tokenholders of Phala (PHA) will receive the new ERC-20 version of the token at a 1:1 ratio, while all current staking, governance, and rewards mechanisms will remain active on the new network.

As per the official announcement on X:

“Proposal Passed: Phala is Migrating to Ethereum L2!” This is a push for greater scalability and enterprise-grade security.”

Why Is Phala Leaving Polkadot?

Unlike other projects in the Polkadot ecosystem that have opted for a multichain strategy, such as Astar and KILT Protocol, Phala is taking the rare step of a complete migration.

What Is Phala Network?

Phala is a decentralized cloud computing network designed to provide private, secure, and scalable computation for AI-integrated Web3 applications.

Initially launched as one of the first Polkadot parachains in 2021, Phala aims to serve as a foundation for AI agents, confidential computing, and enterprise-grade decentralized compute.

Its switch to Ethereum L2 is part of a broader strategy to solidify its position in the growing AI and Web3 compute space.

The Problem With Staying On Polkadot

A key point made in the proposal was the limited scalability of Polkadot’s parachain architecture.

Renewing the parachain slot would lock Phala into infrastructure that requires ongoing resource-intensive maintenance and limits potential growth in enterprise and AI-focused markets.

Ethereum L2 Offers The Edge

Phala already operates a live and functional L2 on Ethereum, launched in January 2025. The migration allows Phala to:

• Tap directly into Ethereum’s deep liquidity and developer tooling

• Reduce operational overhead

• Seamlessly deploy GPU-based confidential compute workloads

• Align with EVM standards, opening access to a broader developer ecosystem

“L2 offers lower operational overhead and direct integration with Ethereum liquidity and tooling,” the proposal reads.

The Vision Ahead: Confidential AI Compute For Web3

The move to Ethereum L2 aligns with Phala’s goal of becoming the go-to infrastructure for confidential AI workloads in Web3.

As part of several key changes to its community, Phala plans to roll out its Trusted Execution Environment (TEE)-based solutions across enterprise clients looking for secure, off-chain compute paired with on-chain verifiability.

Source: X (@PhalaNetwork)

Commercial Traction Is Growing

The proposal hints that Phala is already seeing early-stage adoption for its confidential compute offerings, particularly those powered by GPUs and AI agents.

Moving to Ethereum L2 accelerates integration with more scalable and enterprise-ready environments.

FAQ

What is Phala?

Phala is a decentralized cloud computing network focused on private, secure, and scalable computation for AI and Web3 applications.

Why is Phala moving away from Polkadot?

Phala cites limited scalability, high operational overhead, and a stronger alignment with Ethereum’s ecosystem and tooling as reasons for the move.

Will I lose my PHA tokens in the migration?

No. Tokenholders will receive ERC-20 versions of PHA at a 1:1 ratio, and staking, rewards, and governance will continue on Ethereum L2.

When will the Phala migration be completed?

The migration is expected to be completed before November 20, 2025.

Is Phala abandoning Web3?

Not at all. Phala remains deeply committed to Web3, with its migration aimed at better serving the confidential AI compute needs of the decentralized web.