Submitted by nirlin, also found by pontifex, ladboy233, ast3ros, and 0xpiken
https://github.com/code-423n4/2023-09-ondo/blob/47d34d6d4a5303af5f46e907ac2292e6a7745f6c/contracts/bridge/SourceBridge.sol#L61-L82
https://github.com/code-423n4/2023-09-ondo/blob/47d34d6d4a5303af5f46e907ac2292e6a7745f6c/contracts/bridge/DestinationBridge.sol#L85-L114
Users with account abstraction wallets have a different address across different chains for same account, so if someone using an account abstraction wallet bridge the asset, assets will be minted to wrong address and lost permanently.
Account abstraction wallets have been on the rise for quite a time now and have a lot of users. See the below image for the figures by safe wallet (one of the account abstraction wallets):

With 4.4 million users and 5.4 billion assets, there is very high risk that safe wallet users will try to bridge the assets and loose them.
Now, look at the codebase and understand how the assets will be lost.
In source bridge in burnAndCallAxelar we construct the payload as follow :
Here, we can see the payload passes msg.sender as receiving an address on the other chain, assuming that the user has the same address across all the EVM chains; which is not the case if user is using the account abstraction wallet.
Then, it calls the following function passing the payload, which then calls the callContract function passing the payload to Axelar Network.
Then, on the destination, any Axelar node can call the execute() function, passing in the payload. The tokens will be minted to an account abstraction wallet address of the source chain, but on destination, the same person will not be the owner of that address; hence, tokens are permanently lost.
Give the user the option to pass in the address the tokens should be minted to on the destination bridge. Pass in the warning for account abstraction wallet holders to not to pass the same wallet. Some wallets may follow a deterministic deployment approach to have same address, but as safe explains, that grants nothing, as each chain has its own different state and opcode differences; so even a deterministic approach may generate different addresses.
Invalid Validation
tom2o17 (Ondo) acknowledged, but disagreed with severity
kirk-baird (judge) decreased severity to Medium and commented:
