Facebook has an extraordinary security infrastructure known as the Facebook Immune System to fight spam and other cyber-scams.
The Facebook Immune System (FIS), has a massive defence network and has considerably reduced spam. It took time for FIS to evolve into an all seeing set of algorithms that monitors every photo posted to the network, every status update and every click made by everyone of it’s users.There are about 25 billion of these “read and write actions” everyday. Facebook’s defence system is one of the largest in existence.
It protects against scams by invoking artificially intelligent software to detect suspicious patterns of behaviour. The system can learn in real time and is able to take action without human intervention.
Some months ago there was a notable attack on Facebook. Several users were duped into copying computer code into their browser’s address bar. The code commandeered the person’s Facebook account and stated sending chat messages to their friends, along with a link where the friends could get their own free ipad. Friends who clicked on the link went to a site that encouraged them to paste the same code into their browsers. These type of attacks can generate millions of messages per minute.
Users are less likely to fall for a similar tactic when using email, because a message would have to sent by a stranger. It’s easier to exploit trust relationships in social networks.
To handle such attacks FIS has generated a signature to differentiate between spam and legitimate messages. The links in spam messages that contain keywords like “free” and the IP addresses of the computers sending the messages.
Spammers use multiple machines for the purpose of switching IP addresses, and use link redirection services to change links in real time. FIS checks to see which messages are marked as spam by users and blocked the messages with similar keywords in the text. The system developed a signature that can detect spam within seconds of an attack.
Any defence based on patterns of known behaviour, FIS is vulnerable to new strategies not known to it. “Socialbots” -software that can pose as a human and control a Facebook account can exploit and elude this defence system. The bots send friend requests to random users. Then they send requests to friends of people they had connected with. The number of accepted requests will increase considerably.
Facebook’s privacy settings allow users to conceal personal information from public view. Since socialbots pose as friends, they are able to extract thousands of email addresses and physical addresses from users’ profiles. This information could be used to launch phishing attacks or aid in identity theft.
Social bots behave differently to people that enter Facebook for the first time because they have no real-world friends to connect with.