Written by Nick Heer.

U.K. Home Office Launches Anti-Encryption ‘No Place to Hide’ Ad Campaign

James Ball, Rolling Stone:

The UK government is set to launch a multi-pronged publicity attack on end-to-end encryption, Rolling Stone has learned. One key objective: mobilizing public opinion against Facebook’s decision to encrypt its Messenger app.

The Home Office has hired the M&C Saatchi advertising agency — a spin-off of Saatchi and Saatchi, which made the “Labour Isn’t Working” election posters, among the most famous in UK political history — to plan the campaign, using public funds.

According to documents reviewed by Rolling Stone, one the activities considered as part of the publicity offensive is a striking stunt — placing an adult and child (both actors) in a glass box, with the adult looking “knowingly” at the child as the glass fades to black. Multiple sources confirmed the campaign was due to start this month, with privacy groups already planning a counter-campaign.

Hannah Bowler of the Drum — a publication I had not heard of until I began researching this story specifically, and which does not seem like the most trustworthy source for original information given that it claims to be the “third-biggest marketing website in the world” but does not have a Wikipedia page — speculates that this could also be intended to counter WhatsApp’s “Message Privately” ads. That seems entirely plausible to me. Whatever the case, it launched today.

The Home Office claims that 14 million fewer reports of possible abuse may be filed every year if unspecified social media companies, which clearly refers to Facebook, enable end-to-end encryption. It cites data from NCMEC in making this assertion. But I looked through that organization’s releases and could not figure out from where the Home Office drew its conclusions. The NCMEC says that, in 2020, it received 21.4 million reports (PDF) from platforms like Facebook and Google. Facebook’s platforms provided 20.3 million of those reports; NCMEC does not publish more granular data for Facebook. Perhaps 14 million of those reports were from Facebook and Instagram direct messages — WhatsApp messages are already encrypted — and the remaining six million came from other sources on Facebook’s platforms, like posts and Facebook Groups. But it is entirely unclear, and the “14 million” number appears nowhere I can find on NCMEC’s website. This may be nitpicking, but I think it is important that if we are using numbers to illustrate the scope of a problem, that they should be right.

Everyone who is reading this with some knowledge of end-to-end encryption is surely thinking the same thing: it is awful to know that encrypted messaging can be used for heinous purposes, but it comes with tremendous security and privacy benefits for the rest of us. But this campaign is clearly not for well-informed people, as Ball reports:

One key slide notes that “most of the public have never heard” of end-to-end encryption – adding that this means “people can be easily swayed” on the issue. The same slide notes that the campaign “must not start a privacy vs safety debate.”

What a cynical viewpoint the Home Office and M&C Saatchi must have. Privacy is absolutely a factor, and the Home Office agrees:

[…] End-to-end encryption is valuable technology designed to keep our data and conversations safe. We are not opposed to end-to-encryption in principle and fully support the importance of strong user privacy. […]

Unfortunately, the Home Office goes on to ask for an untenable compromise position. The closest we have seen to a middle ground is Apple’s on-device detection of child abuse materials destined for iCloud — and the backlash was so striking that those plans have been indefinitely delayed.

This is a hard problem to contend with, but the solution cannot be to ban anything that does not leave a trail of evidence, as though such an effort would be possible. The rest of us do not want the GCHQ spying on our messages. Besides, it is not as though law enforcement is actually as “in the dark” as they like to claim.