A historian and military expert says his Google accounts have been locked after the company’s automated content moderation systems falsely flagged his research files as “terrorism-related content.” It’s just the latest example of how Big Tech’s automated content moderation and copyright systems routinely result in major headaches for everyday users.
Ed Francis studies the evolution of military technology over at his YouTube channel, Armoured Archives. But this week, Francis says five years’ worth of research stored on Google Drive has become inaccessible thanks to Google’s automated error.
Francis says the file in question was simply a collection of data on various tanks for a coming video on how military vehicles have evolved across historical conflicts. But Google’s automated systems deemed the file a terrorist threat, resulting in a complete lockdown of his YouTube, GMail, and Google Drive accounts.
Google’s terms of service prohibit the sharing of any content that recruits for terrorist organizations, incites violence, glorifies terrorist attacks, or promotes acts of terrorism. But Francis notes his files are historical, and his content has long been ideologically neutral.
“These files contained no graphic images, no promotion of terrorism, no glorification or recruitment or anything else of the sort,” Francis wrote in his complaint to Google. “There is the possibility that ISIS and so on painted slogans or daubed their letters on vehicles as this is common practice in that conflict.”
Google did not respond to a Motherboard request for comment. After an outpouring of criticism, Google overturned the decision Thursday. In a video addressing the ban, Francis said he lost access to “five years of archival research … including thousands of hours spent in archives gathering information, including my book manuscript.”
Videos by VICE
“Without your help to keep the topic trending, Google would have terminated us,” he added.
Francis’ complaint isn’t an isolated one. Google Drive bugs have sometimes resulted in users losing access to their files for absolutely no reason. But other military researchers say they’ve also lost access to important files falsely flagged by Google systems as a terrorist threat.
The owners and operators of the online Tank Encyclopedia also found their Google accounts locked by Google last year for sharing “terrorism-related content” that was, again, just historical documentation of military hardware and global conflicts. While the lockdowns are sometimes reversed once humans get involved, getting to that point can be arduous.
While there has been an incessant call for “Big Tech” to do a better job at managing problematic content online, such companies have found it largely impossible to police the massive troves of data their systems process everyday. Given the immense scale of data being online across various services, it’s impossible to hire enough human beings to review everything.
Enter automated systems that repeatedly and routinely flag innocuous content as problematic or illegal. That’s been a particular problem with Google’s automated copyright system Content ID, which has been consistently criticized for falsely flagging all manner of legal online content—including content in the public domain—for copyright violations.
Consumer groups argue that if companies like Google and Facebook can’t function properly at the scale they operate at, they shouldn’t be allowed to reach that scale in the first place.
That same problem of scale also applies to customer service. Hacked Facebook users say it’s often impossible to get customer service to let them back into their accounts. Similarly, Google users say getting a hold of a responsible human being can be a headache.
“This is not the first time that sites or pages that study the subject matter have come to grief with Google and although there is a two-day wait for a response it has always been upheld that somehow its promoting terrorism,” Francis said.
