When you think of AI today, you probably think of models that can do it all. They write, they trade, they detect cancer. They also increasingly belong to someone else.

Your code, model, weights, fine-tuned edges of your insights smoothed into someone else’s pipeline. Beyond just centralization of infrastructure, it is the quiet & steady absorption of your intellectual property.

What if you no longer wanted to play that ‘game’?

At CrunchDAO, we are building a world where machine learning belongs to its makers. Where predictive intelligence is not centralized with a few institutions but distributed among the world’s global ML talent.

That’s why we are excited to announce our partnership with Phala Network, a leader in confidential computing for decentralized infrastructure.

Phala brings the missing piece: decentralized TEEs. Bringing additional, advanced privacy protections and verifiability to CrunchDAO’s decentralized modeling pipeline, reinforcing our mission to build a trustless and privacy-preserving predictive intelligence network.

Enhancing Privacy and Verifiability in AI

With Phala’s Trusted Execution Environment (TEE) infrastructure, a model built on a cruncher’s machine is encrypted, transmitted and executed in a confidential container. No one including the coordinators running the competition (crunches) can access the source code or logic inside. 

Crunchers can continue to submit proprietary models and inference data knowing it stays encrypted, private, and tamper-proof throughout our ML network.

How It Works:

1. Crunchers build their model locally without sharing the source code.
2. Crunchers publish meta data on the blockchain (hardware specs, etc)
3. Crunchers securely transmit the model via end-to-end authenticated RA-TLS
4. The model runs inside a Confidential Virtual Machine (CVMs) using a model runner container. 
5. All memory and disk are encrypted; ephemeral keys ensure nothing leaks.
6. The model runner image is open source and publicly auditable. 
7. Remote attestation proves that the cruncher’s model ran exactly as intended.

This is model execution without exposure, ensuring fully private model submissions, ensuring creators maintain ownership and control while keeping models auditable and verifiable. Unlocking a whole new world for anonymous contributors, proprietary models, and privacy-preserving ML challenges.

Key Benefits of the Partnership:

• Private Data: All data remains encrypted during submission and evaluation, meeting the highest privacy standards.
• Private Models: Model creators maintain full ownership, with proprietary code protected at every stage.
• Verifiability: Models and results stay auditable and tamper-proof, empowering decentralized ownership across the CrunchDAO ecosystem.

This partnership advances CrunchDAO’s position at the forefront of confidential AI, delivering scalable, secure predictive intelligence for TradFi, DeFi, and the future of on-chain economies. Together, CrunchDAO and Phala Network are setting new standards for secure, decentralized AI.

About CrunchDAO

CrunchDAO is building predictive intelligence by bringing together over 8,000 data scientists and ML engineers worldwide including 1,200 PhDs. Through structured “Crunches” (modeling challenges), the Solana-based protocol secure model crowdsourcing to build superior modeling capacity compared to traditional in-house teams.

About Phala Network

Phala Network is a next-generation cloud platform providing cost-effective and trustless computing for developers. Using Trusted Execution Environments (TEEs), blockchain, Multi-Party Computation (MPC), and Zero-Knowledge Proof (ZKP), Phala delivers flexible, open-source confidential computing for any application.