Horizen Labs

It's official. $ZEN staking is coming back!

If you've built an AI agent that actually performs, your track record should be earning you money. We're building infra at @AHorizenlabs that rewards agent devs based on verified results - not hype. Running 15-min research calls with builders this month. I'll share back what I learn (anonymized). DM me. #AIAgents #Web3AI #CryptoAI #BuildInPublic #zkProofs

📂 Your DeFi Activity ┃ ┣ 📂 What You Think Is Private ┃ ┣ Your Trades ┃ ┣ Your Strategy ┃ ┗ Your Wallet Balance ┃ ┣ 📂 Who Can Actually See It ┃ ┣ MEV Bots ┃ ┣ Your Competitors ┃ ┗ That One Exchange Employee ┃ ┣ 📂 What Happened Next ┃ ┣ Front-Ran ┃ ┣ Sandwich Attacked ┃ ┗ Copied Trade For Trade ┃ ┗ 📂 The Fix ┣ Confidential Compute ┣ ZK Proofs ┗ Horizen

6/ This is the next step toward a blockchain future that is not just open, but usable for serious financial applications. Read the full announcement on our blog: bit.ly/4lwqK9R

5/ Vela + @horizenglobal makes Horizen the first chain on @Base with native confidential execution, extending Base’s scalable, composable infrastructure into institutional and privacy-sensitive use cases without fragmenting liquidity.

1/ We're excited to announce Vela: a confidential compute layer built to bring private execution to public blockchains. Institutions shouldn’t have to choose between access to public-chain liquidity and protecting sensitive strategies, client data, or trading activity.

If you're building in this space or navigating what's coming from regulators, read our latest piece about what it takes to close the gap 👇 bit.ly/4sa1Qz5

Privacy and compliance aren’t opposing forces. They're complementary. We broke down the real gap in today's on-chain privacy tools, and introduced Ancestral Commitment Compliance, a new framework that lets compliance state travel with funds, update dynamically, and preserve user privacy throughout.

Institutions in crypto have crossed the point of no return. PwC's 2026 report states that major financial institutions are integrating RWAs, and regulators are tightening frameworks across stablecoins, collateral, and DeFi. But everything onchain is still public.

🇮🇳 India has some of the fastest growing ZK developer communities in the world. So we’re meeting builders in person. Zero Doubt | zkVerify India is a city meetup series focused on one thing: proof verification infrastructure. Workshops. Builder showcases. Mini ideathons. Join us in: 📍 Bengaluru — luma.com/xi18xjob 📍 Hyderabad — luma.com/v5hd60ys 📍 New Delhi — luma.com/q4fjq3pe 📍 Mumbai — luma.com/cmxbxlej City by city. Zero Doubt.

Open-source agents are getting better at doing work. The next step is making that work verifiable. • Did the agent follow the right workflow? • Did the output pass the required checks? • Did the result come from an approved process? Kurier helps teams verify proofs through a simple API, so agent workflows can be checked instead of blindly trusted. That’s where this is going↴ less “trust the output” more “verify what happened”

India is building serious zero-knowledge talent 🇮🇳 We’re meeting it in person 🤝 Zero Doubt | zkVerify India is a city meetup and workshop series built for developers who care about verification, privacy tech, and real infrastructure. 📍 Bengaluru 📍 Hyderabad 📍 New Delhi 📍 Mumbai Workshops. Talks. Ideathon. Because adoption doesn’t scale through posts. It scales through builders.

“Every image must be described, labelled, and quality assured.” Meta's smart glasses record people in private moments, like bathrooms and bedrooms. An investigation found that offshore Meta workers were asked to analyze these videos; the users had no idea.

The future of AI isn't just about making models smarter. It's about making them trustworthy, in a way that can actually be proven. We analyzed 13 cutting-edge research papers to map where we are and where this is heading. Read now: 👇 bit.ly/4b7LJeb

5/ Blockchain acts as a public, tamper-proof audit log. And with zkVerify @zkvprotocol, verifying those proofs costs as little as $0.0003. That's not a typo.

🧵 1/ AI is everywhere. But can you actually trust how it was trained? AI models learn from data. But sometimes, that data is sensitive (medical records, financial history, personal messages etc.) So companies use a technique called Federated Learning: your data never leaves your device. Only the "lessons learned" get shared. Privacy preserved. But here is a catch.