Twitter has restricted the ability of attackers to post tweets and perform other actions on behalf of many users who have phone numbers associated with their accounts, but some users need to enable a PIN option in order to be protected.
On Monday, a developer and security researcher named Jonathan Rudenberg reported that attackers can abuse the Twitter accounts of users who added their phone numbers to their profiles in order to use the service via SMS (Short Message Service).
Twitter allows users to control their accounts by sending commands via text messages to phone numbers set up by the company. The supported commands include: following and unfollowing users; tweeting; retweeting; sending direct messages; modifying profile information like name, bio or URL; and more.
The problem is that the origin of text messages can be spoofed and there are services that allow users to do this easily.
"Like email, the originating address of a SMS cannot be trusted. Many SMS gateways allow the originating address of a message to be set to an arbitrary identifier, including someone else's number," Rudenberg said.
This means that if an attacker knows the phone number of a Twitter user and that user associated his phone number with his account, the attacker can issue SMS commands on behalf of the user without actually having access to his phone.
In order to use Twitter via SMS users have to first associate their own phone numbers with their accounts. This can be done from the account settings menu in the Twitter website or by sending a sequence of commands, including the user name and password, to one of three Twitter phone numbers registered in the U.K., Germany and Finland.
Once this is done, users can start sending SMS commands from their phones to Twitter "short codes" -- special short numbers registered by Twitter with mobile operators in various countries -- or long codes, Twitter's internationally accessible phone numbers from U.K., Germany and Finland.
Sending text messages with a spoofed origin address to short codes is not possible, because those special phone numbers are only accessible from inside the operator's network and messages don't pass through an external SMS gateway that could allow spoofing. However, in Twitter's case attackers had the option of sending commands through the long codes.
Facebook and Venmo, a payment service, were vulnerable to the same type of attack, but fixed the problem after being contacted in August and November respectively, Rudenberg said. However, after initially responding to the private report in August, Twitter stopped answering, he said.
As a result, the researcher decided to make the issue public a few months later.
"Our team has already addressed this vulnerability," a Twitter representative said Tuesday via email.
Sign up for CIO Asia eNewsletters.