The government isn’t doing a good job of checking in on its social media screening projects, according to a new report from the Department of Homeland Security’s Inspector General. The report looks at three recent pilot programs for social media screenings of visitors to the US, including a mix of automated tools and manual screenings. The Inspector General hoped to assess whether the programs were accurate enough to continue, but found that the department had set few criteria for how accurate the screenings needed to be. As a result, the report recommends more explicit goals for the future.
One of the projects named in the report ran social media handles collected by Customs and Immigration Services through a system developed by DARPA, described as a “social media screening platform.” The name and details of the DARPA system are redacted from the report, but it appears similar to the previously announced Social Media in Strategic Communication (SMISC) system.
According to the report, Immigrations and Customs Enforcement also used an automated tool to scan for social media accounts from non-immigrant visa holders.
The systems were looking to connect immigration applicants to specific social media profiles, presumably so that any information on social media accounts could be considered as part of the application itself. But it’s unclear how effective USCIS’s tool was at finding the right accounts. The report lists a number of scans that returned ambiguous or negative results, although the specific figures are redacted.
Social media checks have grown increasingly controversial amid President Trump’s aggressive immigration policies, which have spurred an increase in Facebook and Twitter checks at the border. Even more extreme measures have been proposed: earlier this year, DHS secretary John Kelly suggested that visitors be required to turn over the passwords for social media accounts. Today, a coalition of civil liberties groups published a letter calling the policy “a direct assault on fundamental rights.”