One year ago, a gunman killed 17 people at a high school in Parkland, Florida.
A commission there released a report last month, which mentioned among other things, many missed warning signs.
Now, some schools across the country are finding their own ways to make sure they don’t miss a thing.
Natasha Chen visited one district in Connecticut, using software that alerts them to troubling social media posts by their students.
Valentine’s day, 2018.
A former Marjory Stoneman Douglas High School student came back to the campus from which he was expelled…and killed 17 students and staff members.
There were warning signs on social media.
Six weeks before the shooting, someone called the FBI “…about the shooter’s gun ownership, desire to kill people, erratic behavior…as well as the potential of him conducting a school shooting.”
The tip never made it to the Miami field office or its agents.
Greg Boulanger, Director of Public Safety at Bristol Public Schools in Connecticut, is not waiting for people to bring in tips.
An algorithm is doing that for him.
The district pays about $2 per student for a service called ‘Social Sentinel’.
The software scans public social media posts within a geographical area, for certain words and phrases…that could indicate violence.
“I’ve had some pretty serious catches where young ladies have made some claims of acts to commit self-harm,” said Boulanger.
Other times, the trigger words bring up posts that are benign…
“I’m going to kill myself if the Patriots win one more Super Bowl,” said Boulanger.
But he says it’s worth it because of close to 20 real situations the district has intercepted since adopting the program five years ago.
“I think it’s a good idea,” said Rebbecca Bailey.
Rebecca Bailey has four children in Bristol Public Schools.
She still remembers the day of the Sandy Hook Elementary School shooting, less than an hour’s drive away.
“Until I found out that it was not around their school and area I couldn’t eat, couldn’t work. I was just really scared,” said Bailey.
Social Sentinel is just one of many strategies of keeping kids safe here…including surveillance cameras and curtains to block an intruder’s view into a classroom.
But the idea of scanning social media has troubled privacy advocates.
In 2017, Facebook and Twitter stopped giving a company called ‘Geofeedia’ access to their data.
The American civil liberties union said surveillance by Geofeedia violated free speech, because the company helped police departments collect information about people’s political activism.
In 2016, Geofeedia’s CEO said the company has policies in place to prevent the inappropriate use of its software.
The founder of Social Sentinel says his company isn’t looking for people’s political activity…
“Whether it’s a political protest or a cupcake festival, our system is agnostic to that. We are just looking for the language of harm,” said Gary Margolis, Founder and CEO of Social Sentinel.
…language Social Sentinel can search for.
“We’re not going out and eaves dropping and looking to say, ‘we don’t like this student. So let’s go look in their account.’,” said Boulanger.
The system does not allow schools to do their own searches.
Instead, the algorithm is scanning the void, flagging cries for help…and potential threats too.
The Director of Public Safety at Bristol Public Schools says the goal is not to get students thrown out.
He says they don’t automatically run to police and turn over what they find.
The priority is to get a school crisis team to help the student.