Monday, June 30, 2014

Facebook, the Select Committee and one in a million.

Unless you've been hiding under a rock, you probably know about Facebook's cavalier disregard of informed consent when experimenting on people. Now this alone should be enough for Governments to investigate and I'm glad to see that this is being raised to the Commons media select committee by Jim Sheridan MP.

What we already know :-

1) Facebook has the means to target individuals through numerous dimensions. 

2) It has passively examined moods. All Facebook users in the 100 most populous US cities were examined between January 2009 and March 2012.  Now, this is perfectly normal market research.

BUT ...

3) It has actively attempted to alter 700,000 users moods without informed consent. I understand Facebook claims that the one line in the Data Use Policy is informed consent but I doubt any reasonable person reading those T&Cs would. That the result was relatively minor does not excuse experimenting on people without informed consent. Oh, and by minor the 'test induced negative emotions in tens of thousands of people'.

4) The experiment was claimed to have been approved by the local institutional review board on the grounds that "Facebook apparently manipulates people's News Feeds all the time". So far, no-one has quite yet explained what this other manipulation includes and what other experiments have been conducted. Is this other manipulation simply A/B testing (providing two different versions of the same page) to improve interaction with the site? Or have they've been trying to deliberately alter the state of mind of other users in other experiments?

5) It has undertaken political experiments. In November, 2010 demonstrated it could get people to cast a vote. Now, that the experiment was encouraging people to vote is something that many would see as positive. But, it should be noted that the experiment shows that Facebook can, if it so chooses, have an impact.

So now we have to ask ourselves whether it is conceptually possible for Facebook to target supporters of one party, encouraging them to vote whilst targeting supporters and friends of another party and using 'mood contagion' to spread feelings of sadness or hopelessness and discouraging a tendency to vote? Certainly, it appears to have the capability but is this really likely to happen?

If you ask me, I'd put the probability of this happening at about one in a million. But then again, if you asked me last week what is the likelihood of Facebook conducting psychological experiments to alter mood on nearly 700,000 people without any clear informed consent then I'd have put this at one in million as well.

Facebook needs to answer some really hard questions not just on informed consent but also what it is capable of doing, what it has done and what it is doing.  Given the response of an investor sarcastically dismissing this experiment as some sort of joke and the dismal excuses that consent was given in the T&Cs then I'm not convinced that Facebook can be trusted to police itself. 

In this instance then either legislation or in the worst case a freedom's forge effort to replace Facebook with a European service might be needed. I hope the Commons media select committee takes up the call and digs deep. Facebook has form here and do we really want to allow this to get out of control again until we need to have another Leveson inquiry? Just imagine if the news media were pulling stunts like this and realise this form of personalised control of news is more invidious.

Of course, they'll probably say 'sorry' for now and how they'll never, ever do it again ... well, of course they won't ... not until the next time.