On
Feb. 5, the short-message-service Internet firm Twitter announced that since
the middle of 2015, it has suspended 125,000 accounts because they appeared to
be promoting terrorism or similar extremist activities. While Twitter has long maintained rules
against such content in tweets, this is the first time they have made public a
specific number of account suspensions connected with terrorism. This move and the associated problem
Twitter is trying to deal with bring up important questions about the ethics of
communications technologies and the way private organizations have displaced national
laws as arbiters of free speech.
Historically,
communications systems rarely arise in discussions of engineering ethics. For example, I doubt that in the 1950s
the Society of Motion Picture Engineers debated the question of screenwriters
who were blacklisted during the McCarthy communism-scare era. The question of a medium's content was
seen to be almost totally distinct from the technology and engineering it used.
But
gradually that has changed as technical, managerial, and censorship roles have
morphed and merged in the strange new cyberspace world of spam, viruses, and
tweets. The problem Twitter faces,
of groups such as ISIS using Internet services to promote and coordinate
terrorist activities, is real.
Syed Rizwan Farook and his wife Tashfeen Malik apparently drew much of
their inspiration for the attack in San Bernadino, California from Internet
sites promoting jihad. Their December 2015 attack killed
fourteen and wounded twenty-two.
Even messages limited to 140 characters can be used to recruit and
coordinate such things, although there is no evidence that Twitter was involved
in that particular incident.
Nevertheless,
Twitter, with only 3,900 employees, faces the daunting task of enforcing its
Twitter Rules on all 300-some million active users every day. Clearly, much of this task involves
technology to sift through the millions of messages pouring through Twitter's
servers. It also involves the
cooperation of groups concerned about terrorism, with which Twitter has teamed
in an effort to find and suspend violators of Twitter's rule against promotion
of terrorism. But it also involves
fundamental questions of free speech—questions that used to be debated mainly
in the halls of legislatures and courts of law, not in the cubicles of software
engineers. Increasingly, it's the
engineers—or people who work closely with them—making the on-the-ground
decisions about who gets to tweet and who gets their beaks clamped shut.
The
fact that Twitter has gone public with a specific number of account closures is
a move apparently designed to send a message to those who would use the service
for nefarious purposes. It also
serves to raise the status of the company in the eyes of those who are worried
about misuse of the Internet for terrorist activities. And it emphasizes the magnitude of the
problem. Suspending accounts can
be compared to a medical test for a serious ailment. If you get too many false positives, you'll be bothering healthy
people with a diagnosis that later has to be reversed. But if you get too many false
negatives, you let people with a serious disease slip through without
treatment, possibly leading to worse results later on. So the challenge for Twitter is to find
accounts that are being used to promote terrorism in some way and suspend only
those, without cutting off people who are not trying to make trouble.
From
a free-speech point of view, these suspensions could be viewed as
censorship. But even the courts
recognize that free speech has limits—the classical example being the lack of a
right to yell "Fire!" in a crowded theater. So Twitter's actions are justifiable on that basis in cases
where the possible harm to others in the form of terrorist activity appears to
outweigh the value of preserving free speech for all Twitter account
holders.
This
is not a critique of Twitter, by any means. They appear to be taking responsibility for a hard job and
doing it as well as they can.
Looming in the background, of course, is the possibility that if a
family of someone killed in a terrorist attack discovers that Twitter accounts
were involved in planning the attack, the firm might get sued. While I'm not aware of any such suits,
such possibilities always have to be considered when you are dealing with a
large-scale operation involving millions of people.
But
I think the most notable thing about this situation is the way that the
practical basis of free speech, in this case anyway, has spread from the legal
system to international private firms where the parties are mostly anonymous
users, largely invisible software engineers, and company policy makers, in
cooperation with various outside agencies who are all selected by Twitter. The legal system hasn't entirely lost
its influence, in that companies such as Twitter are still responsive to
sustained large-scale legal challenges.
But in the wild-West environment of the Internet, such challenges are
unusual and often politically inspired.
Preventing terrorism is a pretty uncontroversial position politically,
and so Twitter doesn't seem too worried that it will get sued by a coalition of
terrorist groups for what it's doing to their accounts. Terrorists have other ways of settling
such disputes, and I hope they don't use them.
It's
a shame that evildoers have bent the Internet to their will to the extent that
firms like Twitter have to spend a lot of time and effort whacking moles, which
in many cases pop up again right away, either on Twitter or on other more
private Internet communications setups.
But doing nothing would be irresponsible. The knowledge that such suspensions can happen is what makes
most Twitter users behave, not so much the actual suspensions, just as the
knowledge that one is liable to get a speeding ticket makes most people obey
speed-limit signs whether or not there is an actual traffic cop in sight. Kudos to Twitter for kicking suspected
terrorists off the telephone wires, so to speak, and let's hope that their very
public stance against such things forces terrorists into corners of the
Internet where it is harder to recruit people to their cause.
By the way, I have begun to do a weekly tweet summarizing
each blog post. My Twitter handle
is @karldstephan, in case you want to follow me there.
