Posted by Guest on February 21, 2018 in Blog

By Laura Neumayer

Social media, paired with discriminatory surveillance, has in recent years compromised a seemingly personal platform for thousands of Americans and immigrants. Last year, the Immigration & Customs Enforcement, or ICE, branch of the Department of Homeland Security piloted a new program for the monitoring of social media for immigrants and naturalized U.S. citizens. The Brennan Center for Justice gathered experts from the ACLUGeorgetown Law18 Million Rising, and the Center for Technology & Democracy on February 20 to discuss the threat to civil rights and liberties of ICE’s proposed next step in social media monitoring.

DHS has a long history of surveillance, especially in the wake of Executive Order 13769—the Muslim Ban. Initially implementing supplementary visa application questions, the Department began inquiring about immigrant social media handles in late 2016, with no detailed definitions or parameters of which social platforms would be the focus (Facebook, twitter, dating applications, etc.). ACLU representative Manar Waheed emphasized the targets of these questions—Muslim, Arab, South Asian, Middle Eastern, Latino, African, African American, female and LGBT communities—and the resulting deportations, denial of entry, and First Amendment rights chill that occur in turn. “The language of their notice uses the words ‘terrorism’ and ‘national security’ that have been used historically in ways that result in discrimination of Muslim, Arab, Middle Eastern and South Asian communities,” remarked Waheed, elaborating further that these terms and lack of parameters “make it really ripe for discrimination.” The similarities in language used between the Muslim Ban and ICE’s “Visa Lifecycle Vetting” are striking, specifically the ambiguity around what “positive contributions” to society, originally glossed over in section 4(a) of Executive Order 13769, mean where immigration law is concerned. “There is no definition anywhere in American immigration law for what it means to positively contribute to society,” remarked Center on Privacy & Technology Executive Director Alvaro Bedoya commenting on the initial Muslim Ban, further elaborating that identical language was re-birthed into the new ICE vetting. “Fast forward 6 months later to ICE procurement documents posted online July 2017. Lo and behold, ICE proposes to create a vetting system that will evaluate an applicant’s ‘probability of positively contributing to society’ ...This language from this defunct Muslim Ban Order is suddenly reborn in this ICE program.”

In a privacy act notice released September 2017, ICE alerted that social media data would be retained post-naturalization for 100 years after date of birth, with information potentially subject to misuse without clarification as to its purpose. The so called “extreme vetting” initiative has been bred from this initial program, with pilots enacted but shrouded in uncertainty. Notably, the pilot does not specify just what exactly is considered a concern to national security, but goes so far as to specify people from “high risk” countries in its parameters. While these countries nor the criteria are specifically identified, Waheed believes that it can be speculated that Arab and/or predominantly Muslim countries are included. This, she goes on, is based on their usual identification by the U.S. government, as well as the fact that when referencing language, ICE documents only ever specify Arabic, further illustrating the populations and countries targeted by this initiative. 

The “Extreme Vetting Initiative” goes a step further to take this social media monitoring inside national borders, performing data collection and retention on green card and current visa holders via machine-learning software. “There is simply no way that this won’t sweep up U.S. citizens and lawful permanent residents as well,” noted Brennan Center’s Rachel Levinson-Waldman. The process is expected to automatically flag at least 10,000 people a year for visa denial. 

Machine learning, however, is not practical for social media monitoring. Algorithms, while good in theory, cannot perform the way ICE might envision when analyzing social media accounts. Center for Democracy & Technology’s Natasha Duarte emphasized the fundamental issue with algorithm analysis of social media: translation.  “They [algorithms] have much lower accuracy for analyzing non-English language text,” she said, elaborating further to say that this inaccuracy would likely hit minority communities the hardest.

The real-world implications of ICE’s “extreme vetting” would impact minority communities, as it has historically. Discriminatory surveillance is not a new practice in U.S. government agencies. “Ideological monitoring,” according to’s Laura Li, conducted by the U.S. government in the name of “national interest” targeted Japanese Americans and long proceeded their unjust removal and internment. American Muslims are constantly needing to prove their “American-ness” and Asian Americans are the most likely demographic grouping to be accused of financial espionage. “For people of color, including Asian Americans and Muslim Americans who may be naturalized citizens or visa holders, we need to continually demonstrate our assimilation and our allegiance to the U.S. to not be surveilled,” explained Li.

Geofeedia, the Boston Police Department’s previously-used software, is a prime example of the dangers of social media monitoring. The program monitored the social media accounts of activists in the area and collected personal data with no link to criminal activity, typically targeting minority communities. While big-name social media platforms pulled the plug on the program, Li highlighted that it serves as a clear indicator of the direction ICE’s vetting will likely go. “ICE’s use of social media monitoring will only serve to further chill free speech, it would target people of color, naturalized citizens and long-time visa holders for deportation, it’s going to label them as enemies of the state.”

The dangers surrounding ICE’s extreme vetting program are evident, especially for minority and immigrant communities within the U.S. regardless of citizenship or visa-status. “What we’re asking is for Congress to get involved in this conversation and to ask DHS some tough questions,” concluded National Hispanic Media Coalition representative Francella Ochillo. “We have to be able to not just talk about these issues in the abstract, but really think about ‘how does this apply if it were me or someone in my district or my community?’” The program has not yet been implemented and has no clear start date. “Press DHS before this program is employed,” said Ochillo. “We can stop this program before it starts.”

Laura Neumayer is a 2018 spring intern at the Arab American Institute.