Social Media Surveillance and Law Enforcement (MATEESCU)
Area 1:
The amount of intelligence being collected from social media is increasing every year, but users of social media may not realize that. Some may not know who can access their information and may have default views set to public on Facebook, for example, which makes all their posts visible to anyone.
Even when one has restricted posts to friends, one must be aware than any of one's friends can be acting as an agent of the police or actually an undercover officer illegally posing as that person.
Most police departments have no policy or training on how social media surveillance can be conducted, and there is no case law to guide these efforts. Since many youth (13-17) are very active on social media, the risk of overcriminalization of youth, especially youth of color, is very real.
DHS has used social media to monitor protests, and this practice brings a harsh light to their respect for the First Amendment.
Area 2:
I have seen demos of these types of tools, which highlighted outlaw biker rallies and road trips and used geo-tagging to anticipate where criminal or other violent events might occur along the route. With a background in military security, I'm aware of OPSEC, and I try to gauge what I post by guessing who could benefit from it or abuse it. One of my groups is only open to vetted members, and while this restricts the size of the group, it increases the quality of the posts and the discussion that occurs.
I think community digital literacy should include how to regularly review your social media accounts to ensure you are not posting information you're not aware of - and others aren't either. I recommend that folks not permit photos to be tagged of them by others without their own approval.
DHS monitoring of #BlackLivesMatter has unsurprisingly been accompanied by the invention of an imaginary 'war on police' and a labeling by many LE agencies and politicians of the movement as 'terrorist,' while ignoring the actions of right-wing militias in domestic terrorism.
Automated tools that generate a threat score should be viewed as just as biased as a human analyst. One of the differences is that we cannot know what factors the automated tools use to generate the score, and it is difficult, if not impossible, to correct inaccurate factors.
The overall question of privacy and what is legally admissible is a ticklish question, and it will only get thornier, the longer its answer is delayed. As the article states, it may require Constitutional modification.
Comments
Post a Comment