About | HeinOnline Law Journal Library | HeinOnline Law Journal Library | HeinOnline

LOG IN
LOG IN


1 Social Media Monitoring in K-12 Schools: Civil and Human Rights Concerns 1 (2019)

handle is hein.brennan/scommtng0001 and id is 1 raw text is: BRENNAN CENTERFOR CENTER D)EMOCRACY CN E & TECHNOLOGY OR STCE Social Media Monitoring in K-12 Schools: Civil and Human Rights Concerns Systems for monitoring students' social media posts are gaining popularity in K-12 education settings.' Some schools and school districts are turning to social media monitoring as a response to the threat of mass shootings.2 Companies are marketing their social media monitoring services with claims that they can identify sexual content and drug and alcohol use; prevent mass violence, self-harm, and bullying; and/or flag students who may be struggling with academic or mental health issues and need help.' There is limited comprehensive data, but available figures, as well as statements from the companies themselves, suggest that spending by U.S. school districts on social media monitoring tools has risen substantially in recent years.4 However, the claims of effectiveness by companies selling these products are largely unproven, and these tools can endanger the very students they are supposed to protect.s Surveilling students' social media activities raises the following serious privacy, free expression, and other civil and human rights concerns that schools, districts, and legislatures should safeguard against.6 Social media monitoring techniques cannot live up to their claims of preventing violence.7 Most social media monitoring software relies on algorithms to analyze posts. These tools are largely experimental and have significant technical limitations and accuracy problems.' Some of these tools rely on a predetermined library of words or phrases that could indicate potential risks of harm.9 However, many words associated with harm (such as bomb or shoot) are extremely common and have meanings that are entirely context-dependent. These types of tools will produce many false positives,10 overwhelm schools with information, and subject far too many students to unnecessary surveillance given their limited efficacy.1 Machine learning, a type of artificial intelligence in which computers learn to identify patterns from examples, may capture slightly more context, but machine learning algorithms are still limited to recognizing patterns in the text that they have been trained on. They do not possess human-like abilities to interpret the meaning or the intent of the speaker.12 Language use evolves quickly on social media-especially among young people-and novel uses of language are particularly likely to trip up an algorithm. Studies show that social media monitoring algorithms tend to err in recognizing humor, sarcasm, and slang, all of which are common on social media.14