AsianFin -- Hyra Network is aiming to build a sovereign AI future, where users own and govern their AI models, data and compute power instead of relying on centralized providers.
In an interview with Barron's China, AsianFin, and ChainDD, strategic partners of the DePIN Expo 2025 held on August 27 and 28, Hydra Network's Head of Fundraising Bui Minh Ngoc, known as Min, noted that "Sovereign AI Future" - as the core of the Hyra network - means that individuals and organizations can truly control, manage and benefit from their AI technologies.
Furthermore, he stated that the greatest challenge in building a decentralized AI network suitable for a diverse global community lies in the adaptability of society and regulations. Technologies can expand rapidly, but the evolving pace of government, policies, and user behaviors varies.
Min believes that Vietnam has tremendous potential in the fields of blockchain, AI and DePIN. In the next one or two years, the AI and DePIN markets in Vietnam will experience explosive growth.
Min is delivering a speech at the DePIN Expo 2025 in Hong Kong.
The following is the Q&A edited for brevity and clarity.
1. The term "Sovereign AI Future" is central to Hyra's mission. What does a “Sovereign AI Future” actually look like for the everyday person using your platform?
Min: A “Sovereign AI Future” means individuals and organizations can truly own, govern, and benefit from their AI. On Hyra, this translates to:
- Users controlling their data, identities, and compute power rather than depending on Big Tech.
- Everyday people can run AI tasks securely on their own devices, contribute idle compute to the network, and receive fair rewards.
- Communities can co-govern the AI ecosystem, ensuring AI serves society—not the other way around.
2. Hyra leverages Zero-Knowledge Proofs (ZKPs) for cryptographic transparency. How does this technology specifically verify that an AI model's output is trustworthy and transparent for users, and why is this level of verifiability critical for the future of AI?
Min: Hyra leverages zero-knowledge proofs (ZKPs) to enable a model operator to prove that an AI model was executed correctly on a given input and produced a specific output—without exposing the model’s internal parameters (such as weights and biases) or the user’s input data.
Before any computation takes place, the model owner publishes a cryptographic commitment—typically a hash of the model’s architecture and parameters. This serves as a unique, immutable fingerprint of the model, often recorded on a blockchain for full auditability and transparency.
When a user submits data for inference, the zkML framework both runs the model and generates a cryptographic proof. This proof provides a mathematical guarantee that every computational step adhered exactly to the committed model, producing the claimed output. Technically, this is achieved by expressing the machine learning model as a verifiable circuit of equations, which the ZKP system can validate.
3.With over 2.7 million users across 205 countries, Hyra has seen remarkable global growth. What do you believe is the primary driver behind this adoption?
Min: Adoption is hard, retention is harder. Our growth comes from delivering real value: Hyra is not just a platform, but an ecosystem where AI and blockchain reinforce each other. We keep human needs at the center—from useful daily applications to financial empowerment—creating trust and long-term engagement.
4. With a presence in 205 countries, what's been the biggest challenge in building a decentralized AI network that works for such a diverse, global community?
Min: The challenge isn’t purely technical—it’s social and regulatory adaptation. Technology can scale fast, but governments, policies, and user behaviors evolve unevenly. Our task is to bridge these gaps while respecting local contexts and ensuring global accessibility.
5. Hyra isn't just about verification; it's also an infrastructure protocol. How does the network incentivize participants to contribute their computational resources? and how does this decentralized approach benefit both AI developers and end-users compared to traditional cloud providers?
Min: With Hyra, any device—from a smartphone to a GPU cluster—can contribute computing power. This creates a DePIN (Decentralized Physical Infrastructure Network) where contributors earn value (tokens, rewards) for their resources.
For developers: Lower cost, censorship resistance, transparent verifiability.
For end-users: Fair pricing, better privacy, and a truly global resource pool.
In short, Hyra democratizes infrastructure that was once monopolized by Big Tech.
6. How do you see Hyra changing the way developers and companies build AI products, especially compared to relying on big tech's centralized cloud infrastructure?
Min: Hyra provides an open, verifiable, and community-owned infrastructure. Developers no longer need to rely on centralized clouds or accept black-box AI. Instead, they can build products where trust, compliance, and transparency are built in from the ground up.
7. How are AI and DePIN developing in Vietnam?
Min: It's very similar in developing countries. We actually support Vietnam. It is still a good market in blockchain, in AI and in DePIN. But we still need to learn and follow big countries. In Vietnam, the market is still very early. I think in just 1 or 2 years, the AI and DePIN market in Vietnam will become very big.
8. Looking beyond 2025, how do you envision Hyra Network fundamentally reshaping industries that are becoming increasingly reliant on AI, such as healthcare, finance, or media, by providing a verifiable and decentralized foundation?
Min: We envision Hyra as the verifiable foundation layer for AI across industries:
Healthcare: Trustworthy diagnostics where both doctors and patients can verify results.
Finance: Transparent, compliant AI decision-making that regulators can audit.
Media: Authentic content verification to fight misinformation.
By combining decentralization with verifiable AI, Hyra ensures AI becomes a trusted partner for society’s most critical systems.
