The Dangers of Deputizing Google to Bust Child Pornographers

By Jack Shafer, Reuters

27 August 14

on’t be evil” — the first sentence of Google’s “Code of Conduct” — has served as the technology company’s corporate motto since its earliest days. But given Google’s role in the arrest late last month of a Houston man on child pornography charges, perhaps we’ve been misreading it. Perhaps the motto is aimed at its customers, as in, “Don’t you be evil or we’ll have you busted.”

Google, obviously, isn’t the first Internet company to alert investigators of a user who might be transmitting or be in possession of child pornography images. Since the late 1990s, the law has required service providers to report apparent violations of child pornography laws. In 2004, for example, AOL provided a tip that resulted in a child pornography conviction. In 2007, Yahoo took similar action that helped earn a child pornography defendant a 16-year sentence. So far, the courts have rejected Fourth Amendment challenges to these prosecutions, and are likely to continue to do so. No credible sources have appeared to denounce the prosecutions as overkill, and I doubt if any will.

The Houston bust, in which John Henry Skillern allegedly sent explicit images of a young girl to a friend via email, comes a year after Google Chief Legal Officer David Drummond renewed his company’s commitment, which he dated to 2006, to expunge child pornography from the Web and identify its traffickers. As the company’s email policies state, “Google has a zero-tolerance policy against child sexual abuse imagery. If we become aware of such content, we will report it to the appropriate authorities. …”

In its efforts, Google has funded groups that search for the images and, with other companies, has built a shared database of digital fingerprints (via “hashing“) of the images. These fingerprints allow Google and other companies, such as Microsoft and Facebook, to “trawl” accounts for apparent violations of the child pornography laws. The hashing technology, it should be noted, is only as reliable as the database. If you were to create new child pornography this afternoon and load it on to the Web, Google’s algorithms would not automatically detect it as child pornography until somebody identified it, fingerprinted it, and fed it to the database.

Google assuages users who worry it might be scouring their Gmail accounts for evidence of other potential crimes. It told Business Insider this week it does not do that — even though its updated-in-April terms of service leave it plenty of latitude to glean whatever it likes from your account. “Our automated systems analyze your content (including emails),” the terms of service state. “This analysis occurs as the content is sent, received, and when it is stored.”

Still, you’re not likely to get turned in by Google or other providers for sharing recipes for meth or bombs. At least not any time soon. Yet the civil-liberties paranoid — and who among us is not one, thanks to the National Security Agency revelations? — frets anyway. It was inevitable that the all-seeing eyes of the search giants, which excel at locating tiny specks of information hidden beneath terabytes of data, would also perfect ways of cataloging forbidden, illegal content and ratting-out those who transmit it.

Luckily for us, child pornography is just about the only information forbidden by our laws (for purposes of brevity, I’m ignoring the legal prohibitions against spilling nuclear secrets and sharing signals intelligence). But as we continue to digitize the most intimate details of our lives and send them to the clouds for storage, we will become ever more vulnerable to the scrutiny of the companies and the governments who control the switches. Should a moral panic generated by another 9/11-style attack, an epidemic, or a natural disaster sweep through our culture, who is to say that the tools being perfected to apprehend aficionados of child pornography won’t be turned on the rest of us? And it’s not just the big data of lives that needs protection. If nothing else, the Snowden revelations have sensitized us to how revealing even the seemingly innocuous meta-data generated phone calls can be.

Like cannibals, murderers, pedophiles and rapists, child pornographers — and customers of child pornography — constitute the worst of the human worst. They are the exemplars of the retrograde. Our natural impulse will always be to use whatever means, legal or technological, to expose and punish such unrepentant deviants.

Today, I’m fine with Web companies using scanning technology to uncover those who trade in child pornography. But the powers conjured up out of universal abhorrence have a way of spinning out of control, leading us to commit immoral acts in our pursuit of morality. It wasn’t that long ago that marrying across racial lines was a crime. Or that homosexual acts were punished by law. Or that pot smokers were jailed for decades. Or property covenants prevented Jews from buying properties. Should the current powers thrown at the child pornographers not be judged to work, it’s inevitable that otherwise rational people will start calling for stronger security measures — maybe banning anonymity and forcing everybody to carry government-issued Internet licenses? — to end the current scourge.

The old legal cliché holds that bad facts make bad law. To that I’d add this codicil: Bad guys make bad law, too.