Proof-of-Inference

Verifying the Integrity of AI Inferences

Developed by Inference Labs, Proof-of-Inference adds a layer of mathematical verification to any AI function using Zero-Knowledge proofs and cryptography. This process provides reliable verification of AI functions while allowing the nature and results of these functions could remain private. The functions of the Proof-of-Inference mechanism can be summed up as follows:

1. The requester submits input to a model of their choice

2. A distributed network of workers run the requester’s selected model against the input data and generate a proof as well as output

3. The proof and output is returned to the requester, and at any time the proof can be used to prove that the model was run correctly on the requester’s input and produced the output that was provided to the requester

With the assurance of Proof-of-Inference verification users (requesters) can remain confident in the integrity of AI models and outputs.