04 July 2010

Profiling

In a plenary session paper at the Victorian Privacy Commissioner's 2010 Watch This Space: Children & Privacy conference [full version PDF] I noted parental use of 'nannycams' and services that claim to provide parents with lists of what minors have been surfing or copies of what the minors have been texting (ie transcripts rather than merely contact numbers).

I was thus interested to see 'Now Parents Can Hire a Hall Monitor for the Web' by Brad Stone in today's New York Times, describing US services - presumably available in Australia for those with the requisite credit card - that profile what minors are doing online by scraping information from blogs, social network services and other parts of the web.
it comes as no surprise that, after years of headlines and horror stories about predators, cyberbullies and other dangers to children online, a crop of subscription services has emerged to help parents monitor their child's activities on social networks.

These start-ups aim to distinguish themselves from the older category of software products like NetNanny. Such products sit on a user's hard drive, primarily to block various Web sites.

The new companies include SafetyWeb, based in Denver; SocialShield, of San Mateo, Calif.; and MyChild, a service of ReputationDefender, in Redwood City, Calif. These services scour the Web to create easily digestible reports for parents of everything a child is doing online.

The companies charge for subscriptions; the lowest costs $10 a month or $100 a year.
Everything? Not quite. Presumably the 'scouring' can be subverted in much the same way that other electronic handcuff services can be subverted.

Stone asks "For harried parents, the question is: Are they worth it?" and responds -
Certainly not for people who are Web-savvy. The services gather data that can be freely collected with a bit of ardent Web searching.

But many parents are overworked and generally overwhelmed by the rapid pace of technological change and the continuing introduction of social Web sites. For these people, a simple Internet cheat sheet on their child — even at $100 a year — could be a useful tool.

I tested two of the monitoring services, SafetyWeb and SocialShield, on myself, various family members and a baby sitter and found the reports to be a bit unpolished. Both start by asking for a few pieces of information about a child, including his or her e-mail address and the family’s physical address. Then they look through various social networks, checking to see where the child has accounts and, where possible, monitoring what the child writes and what others write about the child.

Long lists of a child'’s online activities emerge, some marked as safe, some as potentially dangerous. Other items are explicitly red-flagged, like a Facebook friend who is considerably older, or a posting with a keyword like "kill" or "suicide".

As you can imagine, there are plenty of reports about innocuous accounts on sites like Amazon and false alarms (“the band killed last night”), for which the companies do not apologize.

"If it's good, we'll tell you about it and if it's something to be concerned about, we will tell you as well", said Geoffrey Arone, chief of SafetyWeb.
That is consistent with disavowals by the service providers of a commercial interest in "stoking exaggerated fear about child safety" in order to sell more subscriptions.
"We are not trying to do any fear-mongering", said Roger Lee, a partner at Battery Ventures, which recently invested in SafetyWeb. "Parents don't need SafetyWeb and anyone else to scare them. They hear about these heartbreaking situations already. We are just trying to give them a product to help solve the problem."
Stone argues that
the services look only for material that is publicly available, which is part of their value: many kids, especially teenagers, need constant reminding that what they post online may be viewed not only by their parents but later by colleges and employers.
There's no indication of whether the services share their surveillance with third parties, a concern given Stone's note that -
Because SocialShield is in the business of finding red flags, it is worth noting that the firm has something of a red flag itself. The company said that one of its consultants is Robert Maynard Jr., a co-founder of LifeLock. Mr. Maynard resigned from LifeLock in 2007, after it became known that he had previously agreed to be banned for life from the credit repair industry amid Federal Trade Commission allegations of deceptive practices at one of his past companies. Arad Rostampour, a co-founder of SocialShield, said that Mr. Maynard was not an investor or a board member but had helped the company to market the service over radio and television.
Are children aware that they are being surveilled? Stone indicates that -
When it comes to Facebook, often the center of children’s online lives, SafetyWeb takes a more discreet and privacy-respecting approach. It asks parents to link their Facebook account to the service, assuming that they are friends with the child on the site. If a parent is not a Facebook friend of the child, SafetyWeb can do little more than record the existence of the child’s account.

By contrast, SocialShield asks the child, not the parent, to link his or her Facebook account to the monitoring service. That gives SocialShield constant access to a child’s Facebook account, even if the child and parent are not friends.

That ends up being a more thorough approach, but it may also be more intrusive. The service collects more information, but the child typically knows about the monitoring.
That knowledge erodes trust, which as indicated in my paper is a foundation of resilience online. It potentially encourages children to evade the surveillance, for example by using a pseudonym and a friend's device, unsurprising given that much of childhood is a process of identifying and testing rules.

The same surveillance models can and of course are being used to monitor the activities of untrusted partners, employees and employers.