Charles Hoskinson announces late-March debut for Midnight, unveils privacy simulation platform

Input Output Global (IOG) founder Charles Hoskinson announced Thursday that Midnight, the company’s long-awaited privacy-focused blockchain, will officially launch during the final week of March.

The announcement came during Hoskinson’s keynote speech at Consensus Hong Kong, marking a major step forward in IOG’s efforts to bring data protection and regulatory compliance to decentralized systems.

“We have some great collaborations to help us run it,” he said. “Google is one of them. Telegram is another. We’re really excited, there’s more that will come.”

Midnight uses zero-knowledge (ZK) proofs to enable selective disclosure. Think of it as a smart curtain for blockchain data, letting users share only what they choose while keeping the rest private. It works as a partner chain to the smart contract platform Cardano and provides privacy and regulatory compliance for decentralized applications.

Alongside the mainnet timeline, Hoskinson unveiled Midnight City Simulation, an interactive platform offering a glimpse of how Midnight’s delivers scalable privacy through selective disclosure. The so-called rational privacy ensures that transaction data remains private by default, while specific information can be shared with authorized parties when required.

This flexibility balances transparency and confidentiality on the blockchain through multiple disclosure views, categorized as public, auditor, and god, each with a different access level.

The simulation, hosted at midnight.city, became operational at 10:00 a.m. Hong Kong time Thursday, although public access to the simulation remains restricted until Feb. 26, according to a press release.

The simulation, which runs on the Midnight network and recruits AI-driven agents that interact unpredictably to create a steady flow of transactions, shows how well the blockchain can handle real-world demand and scales accordingly.

IOG said this test demonstrates the network’s ability to keep generating and processing proofs at scale — an important step in proving it’s ready for real-world use.