There are a few potential issues with physical tokens.
. The entry of the number is eavesdropped on (normally done when a computer is infected with certain malware). [Shared with SMS]
. The user writes their pin number (or other required number) on the token.
With PHONES:
.
I believe that the tokens do not really suffer form being eavesdropped on as they are valid for much less time than an SMS code is so even if someone has the code you used it will have likely expired once they have it/try to use it
Plus once you have used your token code it will not be accepted for another login even if it could still be valid.
The same appears to be true for citibank SMS so as long as you login as soon as you have typed the code in, if your computer is compromised the token code or sms code is of no use to anyone.
If you were to walk away after typing the code in without logging in then someone watching your computer through malware could potentially use the sms code as it is valid for 8 minutes.
There really is not any room to write a username/password on the typical keyring tokens issued however it is possible to store all your account information in plain text in a mobile phone.
(and tokens are expensive, likely around $10, but I'm not privy to corporate pricing)
I have heard of prices much more expensive than that.
$25-$100 is the usual prise bracket that I have heard. However, i do find it hard to see big banks spending $25 on a token for each customer.
All valid points. On balance I still reckon SMS is slightly safer though, given the more likely risky-scenarios are all on the token side of the equation - i.e. likely to take longer to notice a loss/theft of token, user writing PIN on token.
But even if you lose your token it has no identifying marks on it, if you lose your phone it could have your user name and password allowing someone to quickly gain access to your account (though, it is probably unlikely)
---
One thing I do find interesting is that in 2011 for CBA/Netbank mobile phone porting fraud (sms security) had the lowest count of fraud however the highest amount of money stolen for all fraud which occured whereas man in the browser attacks (used with tokens) had the highest count of fraud however the actual amount of money taken was relatively low.
This was in part due to how netbank used the token code as once it was entered anything could be done in the account without further verification whereas with sms security each transaction needs a unique sms code. I don't know if this has since been changed as i am not a netbank customer.
The current citibank system is much like the combank token system in that one code can be used for many transactions.
Last edited: