FBI seeks tools to monitor social media, 'detect mass shooters before they strike'
FBI seeks tools to monitor social media, 'detect mass shooters before they strike'
by Leandra Bernstein
Tuesday, August 6th 2019
WASHINGTON (Sinclair Broadcast Group) — The question is asked in the aftermath of every mass shooting: Why didn't anyone see this coming?
In Texas, the 21-year-old suspected gunman charged with killing 22 at an El Paso Walmart Saturday reportedly drafted a lengthy anti-immigrant manifesto describing his plans and political motivation for the attack and posted it online about 20 minutes before the incident.
In Dayton, Ohio, the 24-year-old gunman who killed nine people, including his sister, was reportedly fascinated with mass shootings, had a "kill list and a rape list" and liked several posts about the El Paso shooter before carrying out his deadly rampage early Sunday morning.
In California, the FBI announced Tuesday that it is investigating the July 29 shooting at the Gilroy Garlic Festival as a domestic terrorism incident after investigators searched the killer's digital media and found he was "exploring violent ideology" and had a target list that included specific churches and government buildings.
And before streaming the mass murder on Facebook Live, the gunman who killed 50 people in a mosque in Christchurch, New Zealand posted his hate-filled manifesto online, along with images of the weapons he used in the attack.
Addressing the nation Monday on the tragedies in Texas and Ohio, President Donald Trump spoke about stopping mass shooting incidents, hate crimes and domestic terrorism before they happen, specifically by targeting "the dark recesses of the internet." Trump then announced that the Justice Department would be working with social media companies and all levels of law enforcement to "develop tools that can detect mass shooters before they strike."
It is not clear what types of new tools the president had in mind. There is virtually no doubt that federal law enforcement already monitors social media with a variety of tools. Law enforcement and U.S. intelligence agencies have long been working with social media companies to stop online radicalization, disrupt propaganda and address other threats. Law enforcement also actively disrupts illicit activity on the dark web.
The FBI recently announced it was looking for a tool to help them scrape social media sites to detect threats from potential terrorists, criminals before they act. In a request for proposal issued last month, the FBI announced it was looking for a "social media early alerting tool" that would allow the agency to "proactively identify and reactively monitor threats" across platforms.
"With increased use of social media platforms by subjects of current FBI investigations and individuals that pose a threat to the United States, it is critical to obtain a service which will allow the FBI to identify relevant information from Twitter, Facebook, Instagram, and other Social media platforms in a timely fashion," the document stated. "Consequently, the FBI needs near real time access to a full range of social media exchanges in order to obtain the most current information available in furtherance of its law enforcement and intelligence missions."
The FBI asks that the social media monitoring tool be able to provide a "full social media profile of persons-of-interest" including "social networks, user IDs, emails, IP addresses and telephone numbers." It must also offer real-time alerts about content based on keywords, subject, geolocation, photo-tagging and other identifiers.
The White House would not confirm or deny whether this was one of the tools the president was referring to when discussing ways to "stop mass murders before they start." The Justice Department did respond to a request for comment.
The FBI has long been working on ways to predict and preempt violence. But there are questions about whether online monitoring could have prevented the mass shootings seen over the past weekend and even more important questions about whether citizens would be asked to give up too many civil liberties in exchange for security.
Rachel Levinson-Waldman, senior counsel to the Brennan Center’s Liberty and National Security Program, argued it is unrealistic to suggest there's a "silver bullet" technology that can predict who will commit a mass shooting or domestic terror attack and such a tool is rife with potential abuse.
"There's no tool out there that can accurately predict these types of events," she stated. "What they're going to do is flag people who arent' actually a risk, they may miss people who are a risk, and they're going to sweep up a lot of innocent people."
Even the best machine learning tools in the private sector are, at best, 70% to 80% effective at identifying threat indicators, she continued. In cases where private companies use algorithms to detect threats, the worst possible consequence is having an account terminated. "When you talk about law enforcement getting into the game, when they have the power of investigation and prosecution...the stakes are incredibly high," Levinson-Waldman advised.
In recent years, there has been some alarm over reports of federal law enforcement keeping tabs on social media accounts. The ACLU recently filed a lawsuit against the Trump administration for it's "extreme vetting" policy that includes surveilling the social media accounts of immigrants and visa applicants.
That particular practice technically began under the Obama administration. Earlier this year, disability advocates raised concerns about the administration's declared policy of monitoring social media accounts for evidence of Social Security disability fraud.
"Law enforcement is very aware of the concerns people have with these tools," said Bobby Chacon, a veteran FBI special agent with expertise in criminal and counterterrorism investigations. "We want to keep every American safe, yet we're aware of the abuses that can take place sometimes when given that kind of power."
Ultimately, it is up to Congress and the public to strike the right balance between what law enforcement wants or needs and what individuals demand in terms of their civil liberties and rights, he added.
From the standpoint of effectiveness, Chacon was more optimistic about the prospects of internet and social media monitoring. He said it is not only possible for law enforcement to develop tools to help predict and preempt violent attacks but that those tools are "imperative."
"Law enforcement has to be conversant in surveillance and getting into all of these spaces," he said, acknowledging the role of online communication over social media and the dark web in radicalization.
He continued that it is possible to predict mass shootings and other incidents with early indicators found online. "Do I think it's possible in 100% of the cases? No."
The FBI has already started to hone in on patterns of behavior exhibited by mass shooters, though the research largely predates the social media era.
In 2018, the agency published a report on the "pre-attack behaviors" of 63 active shooters between 2000 and 2013. Their hope was to educate law enforcement, family, friends, mental health experts, teachers, employers and others on potential early warning signs of violent behavior.
The FBI openly acknowledged that "there is no one 'profile' of an active shooter." Their data showed few demographic patterns among shooters. Almost all shooters were male and more than half were white, but there was no clear trend in ages, educational backgrounds, employment history or even criminal background. Less than one-fifth of the active shooters surveyed had a prior criminal conviction.
https://kfoxtv.com/resources/media/0...?1565128026834FBI study: Pre-Attack Behaviors of Active Shooters in the United States Between 200 and 2013. (FBI.gov)
Where the patterns began to emerge were in their observable behaviors. According to the study, each of the shooters displayed four or five concerning behaviors, observable by others, leading up to the attack. In more than half of the cases, shooters displayed a noticeable change in mental health, concerning changes in interpersonal interactions and some form of "leakage" of a threat or intent to commit a violent act.
All of the individuals surveyed either lived with someone or had significant in-person or online social interactions. In cases where an individual confronted the active shooter with concerns about their behavior, fewer than half wound up reporting those concerns to law enforcement.
The agency concluded, "There is no single warning sign, checklist, or algorithm for assessing behaviors that identifies a prospective active shooter." The FBI also acknowledged that not every concerning behavior warrants intervention and some warning signs may only be clear in retrospect. Still, the agency encouraged careful consideration of red flags and concerning behaviors.
There are currently 15 states and the District of Columbia with so-called "red flag laws" that allow for families or law enforcement to petition a court to temporarily remove an individual's right to have a firearm if they're deemed a danger to themselves or others. Research suggests that dozens of potential violent attacks were likely thwarted when these laws were applied to deny an individual access to a gun.
President Trump has called for expansion of red flag laws or extreme risk protection orders.
Several lawmakers have proposed legislation to encourage states to adopt these laws.
https://kfoxtv.com/news/nation-world...re-they-strike
