Predictive technology is exploding. The arrival of Big Data initiatives by government, as well as a massive industry of data brokers is not only putting privacy at risk, but is offering those with access to the information unprecedented ways to micromanage our lives.
Over the last few years, I’ve covered the arrival of full-on pre-crime police tech (both public and secret) that is beginning to target potential hotspots for crime, as well as to put certain individuals on lists created by artificial intelligence to identify supposed future criminals.
I’ve also covered trends in predictive medicine that have controversially strayed into the field of mental health, painting an algorithmic picture of people who are prime candidates for some form of potential intervention. These measures are often touted as both humanitarian as well as economically prudent, as health professionals and insurers are able to utilize the near constant stream of information provided by Big Data.
But predictive child abuse?
I suppose we shouldn’t be shocked when we begin accelerating out of control down the slippery slope. The UK just happens to be one of the world leaders at the moment for greasing the skids. According to a new report by The Guardian, we clearly see how a variety of social concerns can combine with the “obligation” of government to cut costs wherever possible. The result could be a truly dystopian scenario of entrusting the welfare of children to computers.
Local authorities – which face spiralling demand and an £800m funding shortfall – are beginning to ask whether big data could help to identify vulnerable children.
Could a computer program flag a problem family, identify a potential victim and prevent another Baby P or Victoria Climbié?
Years ago, such questions would have been the stuff of science fiction; now they are the stuff of science fact.
Bristol is one place experimenting with these new capabilities, and grappling with the moral and ethical questions that come with them.
We can also see how parallel trends in data collection can quickly spill over into every area of human life. This is a principle reason why Activist Post has been harping on this topic for a decade. The Guardian specifically equates (in theory of course) consumer behavior with flagging potential victims for abuse.
Machine learning systems built to mine massive amounts of personal data have long been used to predict customer behaviour in the private sector.
Computer programs assess how likely we are to default on a loan, or how much risk we pose to an insurance provider.
Designers of a predictive model have to identify an “outcome variable”, which indicates the presence of the factor they are trying to predict.
For child safeguarding, that might be a child entering the care system.
They then attempt to identify characteristics commonly found in children who enter the care system. Once these have been identified, the model can be run against large datasets to find other individuals who share the same characteristics.
Now, it’s beyond the scope of this article, but child protection services themselves are vastly documented to put children at risk for abuse. So the idea that data extrapolated from that flawed system should inform the wider world about the people who wind up there seems like a very bad idea. I offer this example to show how supposedly accurate computer systems can wind up with bad data from which to draw their conclusions.
In fact, this is precisely the concern that has been voiced by experts in the field who have urged caution when applying algorithms to social systems rather than, say, raw data like consumer purchases and other far more easily tracked metrics. (Please see: “Predictive Algorithms Are No Better At Telling The Future Than A Crystal Ball”). Nearly every application of predictive analytics I’ve come across performs somewhere in the 80% range. I’m sure that the other 20% of families flagged as potential child abusers — and are then dragged through hell because of it — will be happy that it’s all for “the greater good.”
Here is what The Guardian uncovered about how technology is supposed to “safeguard” children.
They include history of domestic abuse, youth offending and truancy.
More surprising indicators such as rent arrears and health data were initially considered but excluded from the final model. In the case of both Thurrock, a council in Essex, and the London borough of Hackney, families can be flagged to social workers as potential candidates for the Troubled Families programme. Through this scheme councils receive grants from central government for helping households with long-term difficulties such as unemployment.
And, by the way, here is the transparency that the public can expect if this is ever fully implemented:
There is no national oversight of predictive analytics systems by central government, resulting in vastly different approaches to transparency by different authorities. Thurrock council published a privacy impact assessment on its website, though it is not clear whether this has raised any real awareness of the scheme among residents.
Hackney council has also published a privacy notice in the past. However, a separate privacy impact assessment, released in response to a freedom of information request earlier this year, stated: “Data subjects will not be informed, informing the data subjects would be likely to prejudice the interventions this project is designed to identify.” Hackney refused to tell the Guardian what datasets it was using for its predictive model.
Those in the U.S. shouldn’t let their guard down on this topic either. In 2016, PBS wrote a very long and informative article, “Can Big Data Save These Children?” where we can see how this already has been pursued by foster care in Pennsylvania.
To keep kids out of foster care, to intervene and shore up families before they collapse, requires the ability to predict a child’s future. A crystal ball. Cherna found his solution in the form of a data model created by an economist half a world away. New Zealand economist Rhema Vaithianathan built two computer models that can predict a child’s likelihood of entering foster care.
One predicts at birth the chances that a child will be abused or neglected and ultimately end up in foster care. The model does this by weighing a number of factors: Among them, does the child live with one or both parents? Has the family undergone prior child welfare investigations? Does the family receive public benefits? The second model uses similar information but activates only when a call is placed about a child’s safety. But – and this is critical – that call often doesn’t come until after the abuse has occurred.
What could go wrong?
“The worst case scenario is that the score is just reflecting the prejudices or beliefs of whoever scored the algorithm,” he said. “It’s very important that these things be transparent so they can be scrutinized by the public and experts.”
But is this really the current path we are on?
Nicholas West writes for Activist Post.