Should Authento DID Adopt ZK?

Apologies, but we don't. We found it currently doesn't meet our service scope or technical needs.

It's been a while since I shared. Recently, my focus has been on Authento, the first startup project.

From my first encounter with @Aom, she brought an idea. Having worked in the compliance field for nearly a decade, she noticed its inadequacies: numerous fake documents, data stored carelessly, users repeatedly asked to submit information, and project parties unfamiliar with compliance requirements. Aom believes user information shouldn't be isolated across projects but should publicly verify results, making the credentials reusable. Thus, users wouldn't need to constantly prove their identity.

This perspective resonated with me: "You don't need to keep proving who you are."

Aom proposed reshaping the compliance field, addressing these scattered information and isolated issues. This is not only user-friendly but also saves projects significant integration and review costs.

After forming the team, various challenges arose: product positioning, structure, technical framework, service scope, revenue model, etc. Among them, the most controversial was: Should we use ZK technology?

Sorry, we tried but eventually dropped it. We found it temporarily couldn't meet our service scope and technical needs.

Undeniably, ZK technology, like AI a few years ago, is worth studying.

The mainstream implementation of ZK is Universal zkSNARK (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge).

There are three ZK solutions, with the latest being Universal zkSNARK. It achieves circuit and parameter separation, allowing trust settings for the same type of circuit to be reused. The advantage of separation is that it significantly reduces trust settings and calculations compared to early zkSNARK.

For Authento, when adopting a technical solution, we mainly consider:

  1. Can it enhance our security level?

  2. Can it meet our desired service efficiency level?

Regarding security:

ZK can help protect user data privacy during external verification. It's commonly understood that you don't need to disclose specific ID information; just inform others that the user is verified. However, for issues like database encryption and secure storage of data sources, ZK can't help us address them.

Regarding efficiency:

  1. First, need to ensure the circuit is simple, like a complex formula requiring more calculations.

  2. Next, consider the calculation parameter size. Facial feature information, for example, is a complex vector or matrix requiring significant computational resources and parallel computing.

  3. Lastly, proof generation and verification. In ZK, generating proof usually requires more computational resources than verifying. In an ideal system, if the number of entities generating proofs is much smaller than those verifying, it's worth a try. However, considering our KYC project, personal information changes, such as ID validity, residence address, and continuous AML updates. This means every KYC entity needs to constantly generate and update proofs, the workload and cost of which far outweigh the benefits brought by the "ZK" concept.

In short, we still can't avoid the original data encryption and secure storage issues, and new problems arise, such as how to optimize computational circuits, improve parallel computing efficiency, and reduce computational costs. So, we put down ZK. It seems we're not following the trend, but we chose a more practical path.

Two years ago, when investing in projects, we chased the Narrative, investing in a "dream," a dream someone would pay for. But when I started the project, I was more concerned about how to realize this dream, how to provide quality service to the first customer, and how to earn the first pot of gold.

Authento won't stop exploring user data protection solutions and won't forget our vision: Users don't need to keep proving who they are, whether in Web3 or Web2. Authento's first stop is challenging the information-isolated Web2; we emphasize user rights and value creation as emerging Web3 players.

____

Redefining Trust and Compliance

Authento is a KYC/KYB/AML solution bridging Web2 and Web3 for seamless compliance in both traditional and decentralized finance.

____

Further Reading:

https://aomthentic.substack.com/p/23e866dd-9728-4782-9406-d32d234e4e3e

https://aomthentic.substack.com/p/164839a2-cf0f-4896-a36f-aefea2e71b44

https://medium.com/manta-network/the-evolution-of-zero-knowledge-proofs-in-web3-b92a68c41c04