Torque Integrates NEAR AI Cloud For Provably Private Assistant Chats
NEAR AI Cloud keeps assistant conversations in hardware-isolated environments with cryptographic attestation, so prompts and financial context stay private.
Today, we’re pleased to announce our NEAR AI Cloud integration, bringing verifiable privacy to the Torque Assistant. Users can research markets, review portfolios, and plan trades through natural language without sending financial context to models that may log, retain, or train on it.
Why privacy matters for financial AI
Most AI assistants ask users to trust privacy policies, a poor fit for finance where conversations include wallet activity, yield positions, borrowing limits, and strategy. One leaked prompt can expose holdings, intent, or compliance-sensitive information.
Torque Assistant helps users act onchain, not create another data exhaust pipe. NEAR AI Cloud lets us offer cloud-scale inference while keeping every chat confidential by default.
What NEAR AI Cloud brings to Torque
NEAR AI Cloud runs inference inside hardware-enforced Trusted Execution Environments (TEEs), including Intel TDX and NVIDIA Confidential Computing. Prompts and responses are processed in sealed, isolated environments that remain inaccessible even to the infrastructure operator.
Each request produces cryptographic attestation, hardware-backed proof of the code and model that served inference, moving privacy from promise to something users and developers can verify.
Benefits for users
- Private by design: Assistant conversations run in confidential compute environments, not on shared infrastructure that treats prompts as training data.
- Financial context stays yours: Portfolio questions, trade ideas, and account-level research stay inside attested enclaves instead of becoming part of a vendor’s long-term data store.
- Verifiable, not just claimed: Attestation gives a cryptographic audit trail for how and where inference ran, aligned with Torque’s emphasis on transparency in finance.
- Same assistant experience: Users keep the familiar Torque Assistant for research, execution guidance, and product help, now with a stronger privacy foundation underneath.
Built for how people actually use Torque
People do not treat financial assistants like generic chatbots. They ask about rebalancing, compare yield opportunities, explore cross-chain moves, and think through risk before signing a transaction. Those prompts are inherently sensitive.
NEAR AI Cloud is already serving privacy-critical workloads across consumer and enterprise products. Torque is applying the same model to everyday onchain finance: powerful AI that respects the boundary between helping a user and owning their data.
Conclusion
Torque exists to make decentralized finance simpler without asking users to compromise on custody or control. Extending that principle to AI means assistant chats should be private, provable, and purpose-built for financial use, not retrofitted after the fact.
This integration is an important step forward. We’re excited to keep building an assistant that helps users move faster onchain while keeping conversations where they belong: with them.
