Lately, a lot of people have been talking about how to bring AI agents into real-world business—and how to move money in and out securely. I happened to come across a post from GOAT Network that focuses on Ziren, the proof engine behind its Bitcoin bridge.
@GOATNetwork aims to build infrastructure for payments and commerce between AI agents. Its security ultimately relies on Bitcoin itself—not just casually calling something a “Bitcoin L2,” but making Bitcoin the actual settlement layer.
Bitcoin itself can’t handle complex transactions, so a bridge is needed. Coins are locked on the Bitcoin chain, while corresponding assets are used on another chain.
The key question is: who decides whether those funds can be withdrawn? GOAT Network uses BitVM3. After a withdrawal, there’s a window of at least one week during which anyone can challenge it. Any dispute is resolved on the Bitcoin chain, rather than by trusting a multisig or committee.
Every withdrawal is backed by a proof showing that the operation followed the rules. That proof is generated by Ziren.
Ziren is a zkVM they built themselves—think of it as a notary machine. You run a program, and it produces a very short proof. Others can quickly verify that it ran correctly without having to run the entire program again.
Why build one from scratch instead of using an existing solution?
Because the bridge’s security is directly tied to this proof. If something goes wrong with the proof, the Bitcoin on the bridge could be at risk. That’s why GOAT Network prioritizes the reliability of its proofs over their speed.
The instruction set is also kept fixed rather than changed casually, so audits and tests from before each upgrade don’t become obsolete.
GOAT Network has also carried out a number of public checks: its paper has been published, it publicly generates proofs for Ethereum blocks on Ethproofs, and it has undergone independent audits, adversarial testing, and a bug bounty program. It has even used Lean to produce machine-checkable proofs. Ziren 2.0 is also moving toward post-quantum security.
For @GOATNetwork, this means that activity on the chain—including payments between AI agents—can ultimately inherit Bitcoin’s security. The bridge doesn’t just move coins; it makes Bitcoin the actual settlement layer.
A bridge’s security doesn’t rely only on its challenge mechanism; it also depends on the reliability of the proof engine itself. GOAT Network built it in-house, takes responsibility for it, and continues to put it through public checks, all to keep this part of the system as controlled as possible. If you’re interested, check out their website and related papers.
@GOATNetwork aims to build infrastructure for payments and commerce between AI agents. Its security ultimately relies on Bitcoin itself—not just casually calling something a “Bitcoin L2,” but making Bitcoin the actual settlement layer.
Bitcoin itself can’t handle complex transactions, so a bridge is needed. Coins are locked on the Bitcoin chain, while corresponding assets are used on another chain.
The key question is: who decides whether those funds can be withdrawn? GOAT Network uses BitVM3. After a withdrawal, there’s a window of at least one week during which anyone can challenge it. Any dispute is resolved on the Bitcoin chain, rather than by trusting a multisig or committee.
Every withdrawal is backed by a proof showing that the operation followed the rules. That proof is generated by Ziren.
Ziren is a zkVM they built themselves—think of it as a notary machine. You run a program, and it produces a very short proof. Others can quickly verify that it ran correctly without having to run the entire program again.
Why build one from scratch instead of using an existing solution?
Because the bridge’s security is directly tied to this proof. If something goes wrong with the proof, the Bitcoin on the bridge could be at risk. That’s why GOAT Network prioritizes the reliability of its proofs over their speed.
The instruction set is also kept fixed rather than changed casually, so audits and tests from before each upgrade don’t become obsolete.
GOAT Network has also carried out a number of public checks: its paper has been published, it publicly generates proofs for Ethereum blocks on Ethproofs, and it has undergone independent audits, adversarial testing, and a bug bounty program. It has even used Lean to produce machine-checkable proofs. Ziren 2.0 is also moving toward post-quantum security.
For @GOATNetwork, this means that activity on the chain—including payments between AI agents—can ultimately inherit Bitcoin’s security. The bridge doesn’t just move coins; it makes Bitcoin the actual settlement layer.
A bridge’s security doesn’t rely only on its challenge mechanism; it also depends on the reliability of the proof engine itself. GOAT Network built it in-house, takes responsibility for it, and continues to put it through public checks, all to keep this part of the system as controlled as possible. If you’re interested, check out their website and related papers.
