With more than 1.6 billion people accessing social media sites globally each and every day, this is a hacker’s paradise. It is bad enough for private individuals who have their social accounts hacked and private information breeched, but now imagine companies that use the marketing power of social networks and you’ve opened up a whole new can of worms. to the integrated Machine Learning (ML) into Strixus® detects potential threats on social sites before they can do any real harm. While Strixus is a cyber-attack prevention platform built especially for large corporations, the technology behind how it works may provide much needed insight for security platforms designed for individuals and small to medium sized businesses.
Where the Threat Is Coming From
Actually, there are so many directions from which hackers are attacking social networks that it’s really difficult to pinpoint where they are coming from. You would probably need to be aware of what it is you are ‘liking’ and links that you click. Perhaps the most widespread attack is coming from the keywords associated with those posts you like or those links you click so this is why it is so important to really understand if your friends and/or followers are really the poster. Hackers take over accounts and then post in the person or business’ name which gets likes and clicks and then has the wherewithal to take over the computer of the person liking/clicking.
Categorizing Social Media Threats
Beyond hijacking your likes in social media, there are actually four main categories of threats which many traditional social media users aren’t aware of. These have been categorized as:
- Link-jacking
- Like-jacking
- Phishing
- Social Spam
While the first two, as described above, are easier to understand and often to spot as well, phishing and social spam can sneak up on you unawares. Phishing seeks to obtain information that is of a sensitive nature such as your passwords, credit card information and yes, sometimes even information about your bank account. Social spam takes on many different forms above and beyond spammy messages that just keep coming and coming. On social networks it often presents itself as a comment from a friend (who has been hacked) and will present itself with such things as hate speech and even reviews of products fraudulently reviewed by a friend – a friend who, of course, has been hacked. So, as you see, threats can come from any direction.
How Strixus Uses ML to Scope out Potential Threats
There is something really beautiful about the combined efforts of Data Mining, Artificial Intelligence, Machine Learning and Human Intelligence. Three of the four are forms of technology while the fourth is, obviously, human. Data Mining gathers large data sets to be analyzed on social sites and then AI seeks to categorize them based on a set of parameters. Strixus has harnessed the power of ML to detect the emotion of the content so that the software is able to scope out whether or not any threat exists based on clues within the text. With much of social media being based on the emotions of the poster, this is truly a huge endeavor and one which, amazingly, Strixus has mastered. If the tone of the content is determined as a possible negative a new set of rules are triggered to determine if it is a threat, without the need of human interference. The content is then queried hundreds if not thousands of rules with this type of logic:
- Is the content published on a known threat location?
- Is the user a known past threat actor or attacker?
- Is the IP of the source in a good or bad repute?
- Does the content link to known threat locations (repeat above)
- Is the location/user/actor on the approved ‘white list’ provided by the client to publish such content
The rules are endless and needless to say, the results are pretty accurate as far as weeding out false positives go.
Because machines are now able to literally scope out the emotions within content, the process becomes infinitely faster than if you had to rely on humans gathering and analyzing data being collected. In fact, it might not even be impossible to do simply because of the speed at which computers can mine data. Through the combined effort of man and machine, social media can be protected against cyber threats efficiently and as time goes on, it just might be that machines can work alone seeking out and destroying threats. In the meantime, man and machine are a winning team and Massive Alliance’s Strixus is now leading in automated cyber intelligence.
