I am always amazed when I see technological innovations that have the potential to reshape the world and impact everyday life. Things like the personal computer, Wi-Fi, the glorious iPhone, self-driving cars, and Artificial Intelligence (AI). We are living in an exciting time where advances in technology have the potential to solve some of the biggest problems we as humans face.
One of these problems, which is more serious now than ever before, is the use of social media to deliver direct threats, such as a student threatening to “shoot up” their school, and also the use of social media to inspire others to commit heinous acts of violence. The incident that occurred this past weekend at Pulse in Orlando, Florida is a prime example, and unfortunately we’re now left with 49 innocent people dead and more injured. Sadly, this is only the beginning.
Our research indicates that social media threats of all types have spiked considerably since 2012 and for the past two years they’ve more than doubled year over year. In 2016, we expect there to be over 2,500 incidents where direct threats were delivered via social media and many more when you factor in propaganda, threatening behavior, etc.
As someone who has been entrenched in social media for the past 15 years when message boards were the talk of the town, and someone who has guided large organizations on complex social media initiatives for over 10 years, the fact that people are losing their lives because of social media is totally unacceptable.
To help solve the problem, those responsible for the safety of others need to be empowered by technologies that allow them to effectively combat the challenges they face whether it’s school threats, workplace violence, terrorist activity or other types of threats.
Law enforcement and intelligence agencies as well as private security teams are stuck between a rock and a hard place: They see the problem, understand the need for improved technologies to combat the issue, but the tools at their disposal simply do not get the job done. I am committed to changing that.
The biggest misconception is that effective social media threat monitoring comes from looking at combinations of keywords, trends under specific hashtags, and other basic analytics that us marketers have been using for many years. Unfortunately, that’s not the case.
In this day and age, it’s imperative to take a proactive approach to identifying social media threats of all types versus a reactive, investigative approach after the fact, and scratching the surface to uncover elementary data points is not going to get us to where we need to be. But by understanding the mindset of threats actors, we can work towards the end goal.
As Soteria Intelligence continues its mission to research and develop innovative technologies that will have a lasting impact on the world, one of our key focuses has been on analyzing preoccupation to understand interests that could elude to intent. Not all signs of distress or impending harm will come in the form of direct threats, and the trail of breadcrumbs leading up to imminent danger are equally as important.
For example, an individual may be innocently standing outside the gate of the White House taking pictures, but he or she could be just as much of a threat as someone who has clearly voiced threats.
Looking at preoccupation, let’s say this particular individual hyperactively engages photos of guns, explosives or violent images on Instagram (or another social network). Such engagement shows a level of interest, and once a certain threshold is reached preoccupation forms.
As a real-world example, Soteria Intelligence was tracking organized crime in San Diego, CA on social networks and in doing so we identified an Instagram account that housed very graphic, violent images with ties to a criminal organization. While analyzing those who have engaged such content and their respective interests/preoccupations, we uncovered one account that stood out from the crowd.
The particular account consistently liked and usually commented on each image within a very short period of time from when they were posted which, in itself, indicates a preoccupation with the content being shared. Interestingly, as the account started sharing photos more frequently, the timeframe within which the content was engaged dropped from 1-2 hours to only minutes in some instances.
We then looked at the types of comments the individual would leave to understand his level of interest and ultimately his mindset. In the end, we formed a connection between the man, who happened to be a Tijuana police officer, and a major criminal organization operating in San Diego.
Equally as important as understanding preoccupation is analyzing friends/followers on social networks to connect the dots and paint a picture of what’s brewing on the horizon.
Looking at ISIS propaganda being distributed through social media, those who have already committed to joining the barbaric movement will likely follow and engage content being disseminated. Additionally, distressed individuals with a propensity for radicalization will grasp onto the propaganda and slowly but surely become radicalized themselves.
Let’s explore two scenarios that show how valuable analyzing connections really is:
- John Doe has a presence on Instagram and Twitter, and on average 70% of the accounts he follows are food-related, 10% sports, 15% travel, and 5% guns. That is not cause for alarm.
- Jane Doe uses the same social networks, however 85% of the people she follows have self-generated or shared/retweeted terrorist propaganda. Of the remaining accounts she follows, 10% post graphic content related to mass shootings and 5% post images of cute puppies.
Clearly Jane Doe has connections that are cause for concern, but by looking at that singular data point it’s unfair to form accusations and imply wrongdoing. On the other hand, that data point combined with many others becomes an integral recipe.
The point is, connections are a clear indicator of potential affiliation and interest, but cannot be used as a deciding factor.
It’s amazing how a word so small can mean so much. Something as simple as intent can mean the difference between first degree murder and second degree murder, and it’s also very applicable in the digital arena.
Uncovering intent is no easy task, requiring an orchestra of data points working together to produce a somber song. But by discovering intent, we have the power (when necessary) to disrupt the period where intentions turn to actions and harm is done.
Aside from the above-mentioned data points which are only a fraction of the equation, we also utilize artificial intelligence and analyze historical data on social media threats to understand probability/trends, which is key.
In a world where social media threat intelligence is based heavily on combinations of keywords that produce false positives as well as subjective analysis of social media behavior where bias is introduced, Soteria Intelligence is excited to be disrupting the industry by taking a data-driven, objective approach to dive much deeper than the tip of the social media threat iceberg.
Those who have lost their lives in senseless attacks will never be forgotten, and their sacrifice is what drives our company to be the best we can be and work to make the world a better place.