Cosmos Co-Founder Claims Hackers Forged Merkle Proof In BSC Cross-Chain Bridge Attack

In a recent share, Ethan Buchman, co-founder of Cosmos expressed his views on the BSC cross-chain bridge attack, saying that the crux of the incident is that hackers can forge Merkle proofs.

In a Twitter thread, Cosmos co-founder Ethan Buchman gave some thoughts on the $100 million BSC hack of Binance on Oct. Binance is the largest user of the Cosmos software.

According to Ethan Buchman, the crux of the hack was that hackers were able to forge Merkle evidence. Meanwhile, this is difficult because Merkle proofs are said to provide high integrity.

Blockchain (and IBC) lightweight clients are built on top of Merkle proofs, and many blockchains store data in Merkle trees so that proofs can be generated that some data is contained in the tree.

The Cosmos chain uses a Merkle tree called IAVL, and the IAVL repository reveals an API that uses “RangeProof”, but it turns out the inner workings of RangeProof are horribly wrong. The problem with IAVL RangeProof’s code is that it allows the Left and Right fields to be filled in the InnerNode, an attacker basically taking advantage of pasting information into the Right field.

This information is never verified and never affects the hash calculation, to make the Validator believe that certain leaf nodes are part of the tree. So they successfully forged the Merkle proof.

Buchman says that while using RangeProof is not a good idea, there might be a way to get around this by pre-rejecting proofs when any internal node fills both Left and Right fields .

For Merkle proofs in IBC, instead of using the built-in RangeProof system for IAVL trees, IBC uses the ICS23 standard to generate and validate Merkle proofs from IAVL trees and the ICS23 code doesn’t have this vulnerability, this explicitly “rejects” RangeProof.

Finally, the Cosmos co-founder introduces a new specification that has been developed according to more stringent processes set forth by the IBC standards. This specification is called ICS23.

It is a common standard for merkle proofs that support many types of merkle trees, including IAVL trees.

“ICS23 followed a more rigorous design process intended to minimize surface area while still being general purpose – a difficult task! As part of this, it explicitly *rejected* range proofs. There are no range proofs in ICS23”

He said

In 2022, the problem of bridge hacking is quite common and complicated with great damage. On February 3, 2022, the Solana-based Wormhole cross-chain protocol was hacked, causing damage of more than $321 million. On March 29, 2022 Axie Infinity’s Ronin Bridge cross-chain bridge with the total damage caused by the attack was over $600 million. On June 24, 2022, Harmony announced an attack on Horizon Bridge with an estimated cost of $100 million. Or recently on May 2 more than 600 million were hacked on Nomad Bridge.

The amount of money for hacking on BSC is not much compared to the chain, but this is also a warning and many lessons need to be learned to increase the security of cross chains.

DISCLAIMER: The Information on this website is provided as general market commentary and does not constitute investment advice. We encourage you to do your own research before investing.

Join us to keep track of news: https://linktr.ee/coincu

Website: coincu.com

Foxy

CoinCu News