Near AI's confidential computing platform lets companies run AI models securely
Near AI has introduced two privacy focused tools that let people use artificial intelligence while keeping control of their information. Near AI Cloud and Near Private Chat handle sensitive data in secure environments, giving users and developers a way to work with AI without giving up their privacy.
Illia Polosukhin, founder and CEO of Near AI, said: “Everyone should own their AI. Today Near AI is taking a major step towards the vision of user-owned AI with the launch of private, verifiable AI products.” He added, “Privacy and verifiability are required to maximize available AI context for users and businesses, with confidentiality and compliance thanks to hardware- and cryptography-based guarantees.”
Near AI Chat
Near AI Cloud gives developers a way to run AI models inside secure hardware environments. It uses Intel TDX and Nvidia confidential computing to isolate each request so that it cannot be accessed from outside.
Near says even it cannot see the data that passes through the system. Every request is paired with a cryptographic proof that shows where and how it ran, and those proofs can be checked independently, giving users real evidence that their information was handled privately.
Near Private Chat is meant to be a simple way to speak to an AI model without creating a data trail. It uses encrypted messages so that financial questions, health concerns, or personal notes stay off the record. Polosukhin said “Everyone deserves access to powerful AI without giving up their privacy. We made Private Chat to prove that’s possible. Literally.”
Near AI Cloud is already being used by a group of early partners, including browser maker Brave which is using it to support its Leo assistant. Brian Bondy, CTO and co-founder of Brave, said: “The only acceptable form of AI at Brave is one that protects privacy to its core. We’re excited to work with Near to offer verifiable privacy and transparency, moving towards giving users unmatched ways to verify that Brave Leo’s operation matches its privacy promises.”
Developers can use Near AI Cloud through the familiar OpenAI API format. Python and TypeScript SDKs handle key management and encryption automatically. Streaming and fine tuning are supported, and the company says overhead is typically around 5 to 10 percent.
Near AI says this launch is only the beginning. Its roadmap includes encrypted memory, AI agents, and decentralized coordination tools built on the Near blockchain. Polosukhin said “Every AI workload -- from a chat session to a national system -- should run privately, provably, and in a way that respects the user.”
What do you think about Near AI's privacy first tools? Let us know in the comments.
