Because of Meta’s use of metadata analysis and proactive account blocking, the company’s plans to expand default end-to-end encryption (E2EE) to all of its messaging services by 2023 will not make child abuse material more difficult to control.
Toby Dagg, acting chief operating officer of the Office of the eSafety Commissioner, told a parliamentary committee that once Facebook’s messaging network adopts E2EE, the US-based “National Center for Missing and Exploited Children (NCME) has declared that it goes dark for their purposes.”
The “vast bulk” of the 29.3 million reports of abhorrent material received by NCME in 2021 “were made via the Facebook chat network,” Dagg testified before an inquiry investigating law enforcement powers in regard to child exploitation.
Assistant Commissioner James Watson of the Australian Border Force also told the panel that “encryption technology” has hampered the agency’s device searches for child abuse material at overseas airports.
“Our border officers are taught to use specific electronic inquiry tools. These gadgets allow us to hold an electronic device till we are happy with its contents “Watson stated.
“We have no particular power to compel the submission of passwords.
“It means our officers may be delayed while we investigate bypass methods that may be able to circumvent the password requirement.”
Nevertheless, Meta’s Australian public policy director, Josh Machin, dismissed the idea that E2EE would impede dangerous content detection.
“Even if we can’t see the content, we’ve been able to conduct some very effective work to analyse the information based on the behaviour of the individuals involved,” Machin told the panel.
“A smart criminal enterprise user behaves significantly differently on encrypted services than an ordinary user.”
Machin stated that Meta focuses on the “removal of false accounts and barring persons from being able to [establish] contact to begin with” using intelligence from “metadata or unencrypted surfaces.”
During Senate estimates last week, she cited a report from her office that evaluated the technologies and procedures of Microsoft, Skype, Apple, Meta, WhatsApp, Snap, and Omegle for banning child abuse material.
According to the study, Meta detected the content using both in-house and off-the-shelf techniques.
Meta employs solutions such as PhotoDNA to identify previously proven child abuse photographs, and Meta is the only platform that employs Google’s Content Safety API and internally developed classifiers to identify new material, but only on messaging without E2EE.
Instagram and Facebook messaging can be configured to be E2EE, and WhatsApp is set to be E2EE by default.
