Solana Foundation launches open-source confidential compute tools for AI
The software uses hardware attestations verified onchain and is currently in beta testing on the Solana devnet.
Key takeaways
- 1Solana Foundation released Solana Confidential Compute for building private AI apps.
- 2The toolset relies on hardware attestations verified onchain.
- 3The software was developed with Nito and is currently in beta on the Solana devnet.
The Solana Foundation has launched Solana Confidential Compute, an open-source toolset designed to enable private artificial intelligence development using hardware attestations verified directly onchain. The new software framework brings confidential data processing features to the Solana blockchain development stack.
The software was developed in collaboration with private artificial intelligence application Nito and is now accessible to developers in a beta release on the Solana devnet. Under the framework, heavy or sensitive artificial intelligence compute tasks run inside secure hardware enclosures, which then output hardware attestations that can be submitted to and verified by smart contracts running on the Solana ledger.
Solana operates as a high-throughput public blockchain platform supporting decentralized protocols and digital token transfers. Because standard transactions on open ledgers are visible to all network participants, privacy tools using secure hardware environments allow developers to protect user data and proprietary models while still relying on the blockchain for immutable verification and consensus.
The launch points to an increasing effort by network maintainers to attract machine learning and artificial intelligence projects requiring confidential execution. Developers will watch how the open-source tools perform during the devnet testing phase before any eventual integration into mainnet applications on Solana.
Reported and edited by the YourKhabri desk under our editorial policy. Nothing here is investment advice.