
On March 15, Tether CEO Paolo Ardoino announced that the company’s AI division will release a “genuine breakthrough” this week, marking an important milestone in its development of decentralized intelligent infrastructure. This announcement comes shortly after the QVAC team launched a new version, which features a completely redesigned user interface and expanded support for Retrieval-Augmented Generation (RAG) document formats.
Tether has been developing QVAC (Quantum Universal Autonomous Computer) for over a year. The platform’s main goal is to enable AI models to run entirely on consumer-grade devices without relying on cloud servers or any API keys. QVAC is part of Tether’s technical division, Tether Data, representing the company—originally known for its USDT stablecoin—shifting toward a broader technological vision.
Ardoino personally demonstrated QVAC live, showing how the system interacts with third-party tools like Asana using the Model Context Protocol (MCP) on a standard GPU laptop, performing complex reasoning and task automation entirely locally, without external computational resources. He described the current stage of Workbench as “very early alpha,” but emphasized that users “can already feel its true potential” before the full release of the QVAC SDK.
Since its first announcement in mid-2025, QVAC has shown continuous development, reflecting Tether’s ongoing commitment in this field:
Behind QVAC’s development is Tether’s substantial financial resources. Public data shows that Tether’s net profit in 2025 exceeded $10 billion, mainly derived from high-yield returns on its holdings of $141 billion in U.S. Treasury bonds. Ardoino has reinvested a significant portion of these excess profits into cutting-edge technologies, including decentralized AI infrastructure, energy, peer-to-peer communication technologies, and more: investing over $200 million to acquire a majority stake in brain-computer interface company Blackrock Neurotech, and investing in Italian robotics startup Generative Bionics. Tether’s EVO division ranked fourth in the global brain-computer interface AI benchmark test in February 2026.
Q: What specific breakthroughs might “genuine breakthrough” involve?
Based on QVAC’s development trajectory, potential breakthroughs include a fully open-source version of the QVAC assistant framework (allowing developers to integrate local AI into their applications); advanced edge-optimized language models for inference; or deeper integration between AI agents and Tether’s payment ecosystem (which already supports autonomous trading of Bitcoin and USDT), enabling AI agents to execute cryptocurrency payments directly.
Q: How does QVAC fundamentally differ from mainstream AI assistants like ChatGPT or Claude?
The key difference lies in the architecture. ChatGPT and Claude rely on cloud servers for inference, requiring user data to be sent to remote servers for processing. QVAC aims to perform all AI inference entirely on the user’s local device, keeping data on the device, without needing API keys or dependence on centralized AI service providers. This approach not only enhances privacy but also aligns with Ardoino’s vision of replacing centralized big tech platforms.
Q: Should we be skeptical of Tether’s investments in AI?
Maintaining caution is reasonable. As the issuer of USDT, Tether has faced scrutiny over its reserve transparency in the past. The announcement of a “breakthrough next week” carries a clear social media hype element. The core issue is that there is often a gap between official announcements and actual product delivery. While QVAC’s progress demonstrates genuine technical advancement, the scale and usability of the claimed breakthroughs will only be clear after actual release.