At a tech industry conference hosted by the Government Technology & Services Coalition last month, Immigration and Customs Enforcement (ICE) invited software providers to begin the process of creating algorithms that would monitor the social media accounts of visa holders deemed to be a high risk in order to assess potential threats to the US. The agency announced that they would need tools equipped with “risk-based matrices” that would continue social media surveillance throughout these visa holders’ stay in the US so that ICE may predict any threats. These requests are the first clear plans showing ICE’s intent to augment tougher visa vetting with the monitoring of social media through a program now named “Visa Lifecycle Vetting.”
At the conference, Louis Rodi, deputy assistant director of ICE Homeland Security Investigations’ National Security Program, explained to company representatives that millions of people enter and depart the US every year and the agency needs to be “smart” about how they monitor visitors. He went on to explain that the monitoring of these targeted visa holders’ social media accounts would be large-scale and continuous. “Everything we’re dealing with is in bulk, so we need batch-vetting capabilities for any of the processes that we have,” Rodi said at the conference. Alysa Erichs, ICE Homeland Security Investigations’ acting deputy association director for information management, told those at the conference that the agency hopes to receive automated notifications from this new software about any visa holders’ social media activity that would “ping us as a potential alert.” Rodi further stated that the agency would also be open to other techniques that would allow them to monitor social media activity, such as link analysis, which would help map out applicants’ online connections as long as they are public posts.
Despite these requests to tech companies, the Department of Homeland Security (DHS) has not begun creating any programs yet, according to ICE spokeswoman Carissa Cutrell. Cutrell tells ProPublica that the “request for information on this initiative was simply that—an opportunity to gather information from industry professionals and other government agencies on current technological capabilities to determine the best way forward.” Cutrell also emphasizes in an email that before any program is implemented, it will need clearance from various DHS units including the Privacy Office and the Principal Legal Advisor. ICE, however, has already begun monitoring some social media networks at eight Homeland Security Investigation posts throughout the world and, according to Rodi, plans to expand to more locations.
Rodi told the audience at the conference that ICE has participated in several meetings with companies regarding this software, but did not name the companies. Top tech companies that attended the conference included Microsoft, Accenture, Deloitte, and Motorola Solutions. When Propublica asked representatives from these companies to comment on their level of interest in the proposed technology for monitoring social media, they either refused to comment or did not respond at all. Microsoft officials have strongly opposed many of Trump’s strict immigration policies and Microsoft researchers have publicly called for ICE to stop “spying” on visitors’ social media accounts. “There’s a growing trend at the Department of Homeland Security to be snooping on the social media of immigrants and foreigners and we think it’s an invasion of privacy and deters from freedom of speech,” Adam Schwartz, an attorney for the Electronic Frontier Foundations, tells BuzzFeed News. Many, like Rachel Levinson-Waldman, senior counsel to Brennan Center’s Liberty and National Security Program, fear that these new tactics proposed by ICE could also affect the rights of US citizens who use social media to communicate with immigrants and foreign nationals by collecting collateral data from them as well.
Many immigration and privacy advocates have expressed concern regarding this new proposed technology and the threats it poses to privacy rights and Constitutional protections. Alvaro Bedoya, the executive director of Georgetown Law’s Center on Privacy & Technology, explains that despite the program’s claims that only approximately 10,000 high-risk visa holders would be monitored, the actual number of visa holders affected would be substantially larger.
A group of approximately fifty tech experts, including computer scientists, engineers, and mathematicians, recently sent the head of DHS a letter expressing their grave concerns over the implementation of social media monitoring. The letter states that a computational program cannot “provide reliable or objective assessments of the traits that ICE seeks to measure.” These experts also believe that this program is likely to be “inaccurate and biased.” Bedoya says: “ICE is building a dangerously broad tool that could be used to justify excluding, or deporting, almost anyone. They are talking about this as a targeted tool, but the numbers tell a different story.”
ICE officials and supporters of the Trump administration’s strict immigration policies believe that gathering social media data from those visa holders deemed to be a threat is necessary. “This administration is big on immigration enforcement, so we’re not going to look the other way like we have in the past when we have overstays,” Louis Rodi says. “Maybe it’s an administrative violation—it’s still a crime. These people need to pay. They can’t get away with it.” So far, monitoring the social media accounts of visa applicants has not identified any potential threats against the US that would not have been likely to appear in existing government databases, according to Rodi; however, Rodi believes that one day social media data may provide valuable information not discoverable via other ways.
In the letter sent to the head of DHS, the fifty signers warn that ICE’s interest in these algorithms to assess risk is misguided, considering how rare it actually is for foreigners to be involved in terrorist attacks. According to the Cato Institute, which the letter cited, the likelihood of a US citizen dying on American soil in a terror attack in any given year was about one in 3.6 million between 1975 and 2015. Cathy O’Neil, one of the signatories and author of “Weapons of Math Destruction,” says that any algorithm a company proposes comes equipped with human calculations. “At the end of the day, someone has to choose a ratio. How many innocent false positives are you going to keep out of the country for each false negative?”