The American Civil Liberties Union recently slammed the Fresno Police Department for testing social media-screening programs, suggesting police could use them to monitor protest groups and accusing the department of keeping the public in the dark about the testing.
But police say they’ve only been testing services for possible use in monitoring violent crime and terrorism – not for spying on critics. They add that the public will get a chance to weigh in when a final recommendation goes before the City Council.
Fresno police last year participated in free trials from the social media monitoring programs Geofeedia, LifeRaft and Media Sonar. The department remains on an extended free trial for Beware, a data-mining program that includes social media and, upon request, assigns a “threat rating” to people and addresses.
Fresno activists alerted ACLU representatives about Beware earlier last year. So the ACLU sent out a Public Records Act request to find out how the Police Department was tracking social media, and got 88 pages of documents in return.
Sign Up and Save
Get six months of free digital access to The Fresno Bee
The records included email exchanges with representatives of the four programs Fresno police confirmed they tested.
Social media surveillance programs generally scan large batches of public posts on networks such as Facebook, Twitter and Instagram. Media Sonar especially worried ACLU technology and civil liberties policy attorney Matt Cagle because it allows police to identify threats to public safety by tracking certain keywords, including “BlackLivesMatter,” “ImUnarmed” and “RIPMikeBrown.”
“This kind of surveillance can place people under suspicion simply for speaking their mind online,” Cagle said in a blog post.
Fresno police Sgt. Steve Casto said the programs are not meant to make people feel like “Big Brother” is watching because it’s all publicly and commercially available data. He said the goal is to monitor things like violent crime and terrorism. Police never searched for “Black Lives Matter”-related keywords, he said.
To monitor protest groups or demonstrations – that’s not our aim.
Sgt. Steve Casto, Fresno Police Department
Casto said social media currently is used only once officers have the name of a suspect and can look them up like anyone else would.
“If someone was threatening to bring a gun to a specific high school or mall, we could do a geofence (using Google maps) and monitor for a gun or mass shooting,” he said. “To use it to monitor protest groups or demonstrations – that’s not our aim.”
Documents the ACLU received do not show Fresno police monitored protest groups. But Cagle noted that Fresno activists protested one year after an officer shot unarmed black teen Michael Brown in Missouri, the same time period as documents show the department received emails from Media Sonar.
Cagle said the bigger issue is that Fresno police officials were not open about how they were using and experimenting with surveillance programs. He said Fresno should establish a policy that ensures there is public discussion before using potentially invasive software.
“The community should have a chance to discuss and weigh the costs to civil liberties and costs to taxpayers before their elected leaders make a decision to move forward or not on these technologies,” he said. “We shouldn’t just be learning about these software programs right now.”
We shouldn’t just be learning about these software programs right now.
Matt Cagle, ACLU attorney
Using Beware, police can search a home address to find people, phone numbers, court convictions, relatives and questionable social media posts. The program uses a mathematical formula to judge the potential threat level posed by people connected with the address and assigns a color rating of green, yellow or red.
During a City Council workshop on the testing in early December, some council members took issue with the color rating and social media aspects of Beware. Fresno police told Beware programmers to remove those parts. When police start using the program again in a couple of weeks, the only threat indicator they will see is if the people living at a given address have a criminal history.
Beware would cost around $20,000 per year, while a social media monitoring program would cost a bit below that. Casto said he expects to recommend discussing the purchase of both with the City Council within a couple of months, at which point the public could also offer comment.
The programs would be implemented into the department’s crime-tracking technology hub, called the Real Time Crime Center.