Cysic and Inference Labs partner to launch scalable infrastructure for verifiable AI applications

Trenton, New Jersey, United States, December 22, 2025, Chainwire

Inference Labs, a developer of verifiable AI stacks, and Cysic, a provider of distributed computing and zero-knowledge (ZK) hardware, have formally announced a strategic partnership to enable scalable, cost-effective verifiable AI applications. This collaboration combines distributed ASIC-powered computing with a verifiable AI framework to address long-standing performance and cost limitations in real-world zkML use cases.

The integration deploys Inference Labs' verifiable AI software on Cysic's distributed computing network and leverages ZK proof hardware to support scalable, low-cost AI verification. This unified infrastructure is intended to support new applications such as verifiable autonomous agents and oracles by enabling accountability and transparency through a proof-based system.

Historically, AI’s “black box” decision-making processes have raised concerns about trust and transparency in critical environments. This partnership aims to provide the foundational infrastructure needed to make verifiable AI systems viable at scale.

The new frontier of verifiable applications

The output of today's AI cannot be verified. End users and intermediate applications have no way of knowing whether a particular output is from a particular model. However, future applications will require an auditable paper trail of all AI output.

for example:

Autonomous agents executing transactions must prove that their actions were correct and as prescribed.
Oracles that provide critical data (prediction markets, loans, insurance, etc.) must prove that the data was processed by a genuine, unmodified model.

Extended application: These applications enable a new generation of high-stakes use cases that were previously not possible, including:

DeFi and on-chain finance: AI-powered trading strategies are proven to be fair and resistant to front-running. An underwriting model that assesses risk without exposing sensitive data.
Governance and DAOs: Fully autonomous DAO executives make and execute decisions with cryptographic proof of compliance with governance rules.
Prediction markets and forecasts: An oracle that provides auditable predictions about events, products, and elections with proof of data source and model integrity.
Identity and reputation: A Sybil-resistant identity verification system that verifies uniqueness without collecting or exposing an individual's biometric data.

Historically, the vision for these verifiable agents and oracles was clear, but the “proof generation” step, the cryptographic workload that enables verification, lacked the infrastructure that could scale with the workloads these applications required.

trusted architect

The new partnership creates a synergistic combination between two protocols that are uniquely positioned to power the verifiable AI of the future.

Inference Labs' Validated AI Stack is an integrity layer for AI. Its zk-proof-based system allows autonomous agents and oracles to generate proof of their honest behavior while keeping their core IP and data private. This validation is a fundamental requirement for future AI-powered software.

Cysic's distributed computing network serves as Web3's proofing engine. Specialized hardware (ASICs and GPUs) is used to perform the intensive computations required for the ZK proof, making the execution of these agents at a scale and cost that is economically viable. Together, the two protocols form the infrastructure needed to power a validated AI revolution now and into the future.

“Cysic was built to transform computing into a verifiable, scalable resource for the next generation of distributed systems,” said Cysic CEO Leo Fan. “Partnering with Inference Labs allows us to apply that vision directly to zkML, removing cost and performance barriers and enabling responsible AI across autonomous agents and decision-making platforms.”

“True innovation happens at the intersection of great teams,” said Ron Chan, co-founder of Inference Labs. “Our partnership with Cysic connects best-in-class verifiable AI software with a best-in-class decentralized proof engine. Together, we provide critical infrastructure that enables builders to create a new generation of trustworthy AI-powered applications.”

This partnership marks the activation of a critical infrastructure layer for a trustless digital future. By solving the fundamental bottleneck in proof generation, Cysic and Inference Labs are not only enabling a new class of applications, but also laying the foundation for a world where autonomous systems operate with cryptographic certainty rather than blind faith. The era of verifiable and accountable AI could begin now.

About Cysic

Cysic is building a verifiable compute engine for Web3, backed by major investors including Polychain Capital, OKX Ventures, and HashKey Capital. By combining custom ZK hardware, a decentralized node network, and a programmable economic layer, Cysic transforms computing into a trustless on-chain resource. This network supports scalable proof generation, AI verification, and scientific computing workloads, laying the foundation for the ComputeFi economy.

Follow X: @cysic_xyz

Read the documentation: https://hackmd.io/@Cysic

https://docs.cysicfoundation.org/~/changes/18

About Inference Labs

Inference Labs builds a decentralized AI infrastructure that uses cryptographic proofs to make AI output verifiable, private, and trusted. The company's Proof of Inference protocol and network enables mathematically auditable AI across autonomous agents and Web3 systems, addressing trust, transparency, and privacy at scale.

Follow X: https://x.com/inference_labs

Read the documentation: https://docs.inferencelabs.com

Media contact:

felix@theprgenius.com

jamie@theprgenius.com