The shame continues

Last Thursday afternoon the LIBE Committee of the European Parliament met. The proceedings are viewable.

The meeting was scheduled to take two hours.  It had several items to discuss. The earlier ones overran. When the question of the derogation and children being sexually abused on the internet was reached the Chair of the Committee opened with the following words

“Let’s see if we can manage in 10/15 minutes the three points left on our agenda”.

The discussion of the derogation began at 15.35.  It ended at 15.51 so it got more than its fair share of what was supposed to be the remaining time but, even so, many will feel after all the energy that has gone into this issue, to get 16 minutes was, well, disappointing. But perhaps no more disappointing than the reasons why the discussion was necessary in the first place.

How did we get here?

Back in 2018 the EU adopted the European Electronic Communications Code (the Code). The law was due to take effect on 20th December, 2020.

In the early part of 2020, Commission officials realised some might interpret its provisions to mean it was no longer lawful for companies to continue voluntarily scanning messaging services looking for child sex abuse content or activity. 

Why? Because it became apparent the Code would be subject to the overarching provisions of the GDPR.  This raised the possibility that end-user consent was therefore required before the relevant data processing could take place. Alternatively, as a condition precedent, to allow the child protection tools to continue being used as before there also needed to be a much more detailed and broader legal framework put in place. 

Both these points are contested, but the key question here is had anyone in the Parliament, Council of Ministers or Commission spotted and considered any of this when the 2018 measure was being drafted, debated and finally adopted? No.

Children were forgotten or overlooked. Out of sight. Out of mind.  Simple as that. And not for the first time.

Did the European Data Protection Supervisor or his predecessor step in at any point during the co-legislative process or very soon after it concluded specifically to draw attention to these matters? No.

If anything, when the Supervisor did make an appearance, he contrived to make things worse by issuing an Opinion in which he omitted  to mention a child’s right to privacy or Article 24 of the EU’s Charter of Fundamental Rights which, just to remind you, reads as follows

  1. Children shall have the right to such protection and care as is necessary for their well-being…..
  2. In all actions relating to children, whether taken by public authorities or private institutions, the child’s best interests must be a primary consideration.  

There are several EU-specific and other legal instruments which likewise make clear children are in a separate and special class, that they do require extra care and attention which would not apply to other classes or groups. This fact appears to have escaped the Supervisor who, far from considering the particular position of children, felt free to say

“The EDPS wishes to underline that the issues at stake are not specific to the fight against child abuse but to any initiative aiming at collaboration of the private sector for law enforcement purposes. If adopted, the Proposal, will inevitably serve as a precedent for future legislation in this field.”

Actually the issues at stake are specific to the fight against child abuse. Children should not be lost in a muddled conflation.

In other words, at the relevant time, all the European Institutions, including the political parties within the Parliament, failed to consider the Code’s impact on children’s rights and welfare.  They also failed to consider the express legal burden placed upon them to take account of the unique position of children. If any “precedent” was to be set it would be one about children. Nothing else. The scrutiny processes which are said to be fundamental to the way new laws take shape in a democracy let children down. Badly.  

Protecting children through scanning was not being done covertly

Please bear in mind none of the companies scanning messaging services to keep children safe made any secret of it. On the contrary they spoke about it openly and often.  They were proud of it and wanted people to know.  They thought it reflected well on them. They were right. It did.

The companies were frequently praised for their efforts by the most senior figures in our European Institutions, including in the Parliament.

In some jurisdictions these facts alone would entitle the companies to invoke the doctrine of estoppel. And if it didn’t do that it would almost certainly provide a strong defence against any or many forms of hostile legal action.

Unintended, unforseen and unmentioned

There is no getting away from the fact, for the reasons just given, this is all a terrible mess. A legislative accident. As far as I am aware absolutely nobody wanted us to be in the position we are now in. When accidents happen people generally pull together, at least to get things back to where they thought they were before.

Not here.

Let’s not forget, when the Commission proposed a solution, the temporary derogation, they were not seeking to close down discussion or debate. They were not seeking to make a final, irrevocable decision. They only wanted to create a breathing space so as not to interrupt or give anyone an excuse to interrupt child protection work which had in some cases been going on since 2009.

But the LIBE Committee would not co-operate. They would not help to put things back where everyone thought they were before.  And they did it in the name of defending the rights of political parties to comment and scrutinise. See above for how well that worked last time out.

The result?

A 46% drop in reports of child sex abuse followed this robust assertion of the rights of political parties.

Two wrongs do not make a right

True enough two wrongs do not make a right. Just because EU institutions botched the process in the run up to the 2018 decision, and since, it is no reason to allow them to botch it again, if that is what you think  happened. But that would only be  true if LIBE was being asked to endorse a permanent or longer term position.  Which they weren’t. They were only being asked to agree to a stopgap.

It is clear why, under existing arrangements,  LIBE took the leading role in shaping the Parliament’s decision, nevertheless they are not the most obvious place one would look  to first when a matter so intimately connected with children’s rights was at issue.

Perhaps the Parliamentary authorities could reflect on that. Could the Intergroup on Children’s Rights be converted into a full Committee of the Parliament in an effort to ensure nothing like this can happen again? This might help guarantee every legislative measure is considered expressly in terms of its impact on children. 

Threat to money: we fight. Threat to kids: we don’t fight 

At the LIBE meeting Sophia IN’T Veld made one of the most telling –  I would say crushing – points when she criticised the transparent hypocrisy of Facebook. 

On 21st December Facebook immediately stopped scanning.  IN ‘T Veld pointed out that on a great many previous occasions, with its enormous war chest and great phalanx of lawyers, Facebook had gone to court at the drop of a hat to contest a point if it looked like it might interfere with their ability to make money.

When it came to defending the good work they had been doing for years to protect children they rolled over without a whimper on Day 1.

Preparing the way for end-to-end encryption

A great many people saw Facebook’s decision to stop scanning as being linked to their larger ambition to introduce end-to-end encryption. If there are no tools which can be legally used to scan, it renders otiose the discussion on encryption.  Despite their protestations to the contrary, by saying they agreed the Code made the use of the tools illegal, Facebook were providing legitimacy, succour and comfort to forces who want to see the tools banned forever. 

Simply to believe Facebook could have made a calculation of that kind tells you a lot about how low people’s opinion of them has sunk. In that sense whether it is actually true or not barely matters. And we still do not know when, or indeed even if the decision to stop scanning was discussed internally at the so-called “Safety Advisory Board”. Maybe they should rename it the “Danger Enabling Board”. Just a thought.

IN ‘T Veld contrasted Facebook’s decision with that of Microsoft, Google, Roblox, Linkedin and Yubo. These companies decided the legal risk was minimal to non-existent so they carried on scanning as before.

Complying with the industry standards we like. Disregarding the others

Appearing recently before British Members of Parliament,  Facebook told the MPs one of the reasons they wanted to proceed with end-to-end encryption is because it is now an “industry standard” . Really? Well what about the industry standard established by Microsoft, Google, Roblox, Linkedin and Yubo?

I would dearly like to know who in Facebook made the decision to stop scanning. One has to imagine it was Zuckerberg personally, and if it was then I would say the prosecution can rest its case.  In his Harvard dorm he had a great idea and manifestly Zuckerberg is talented at making decisions which generate tons of cash. But he is plainly not fit to run one of the world’s largest and most important companies as a personal project, empowered as he is by his majority shareholding. 

Posted in Child abuse images, Consent, Facebook, Google, Internet governance, Microsoft, Privacy, Regulation, Self-regulation, Uncategorized | Leave a comment

Canadians, ethics and porn

A great many people in Canada are extremely angry and more than a little embarrassed at the fact that they provide a home and operational base for MindGeek, owners of the world’s largest pornography web sites, in particular Pornhub. That anger rose to a crescendo following the publication of a major article in the New York Times in December. The title of the article was “The Children of Pornhub”.

Self-evidently the good northern folk are not standing idly by gazing into the middle distance, twiddling their thumbs. They are now scrutinising MindGeek from a variety of different angles. And if you need reminding why, just listen to and watch the testimony of two people who appeared earlier this week before the Canadian Parliament’s Standing Committee on Access to Information, Privacy and Ethics. The topic of the hearing? “Protection of Privacy and Reputation on Platforms such as Pornhub”. Roll forward to around 12:28.

The hearing opens with the harrowing evidence of an exceptionally brave young victim, Serena Fleites. Her case was prominently set out in the New York Times piece.

Now 19 years old, when Serena was 14 the person she thought of as her (first ever) boyfriend bullied her into making and sending a sexual video. Then he betrayed her. The video ended up going all around the school, neighbouring schools, and finally on to Pornhub. This was unquestionably an illegal video, containing child sex abuse material. Its publication had catastrophic consequences for Serena. But she did not retreat into the shadows. She fought back. Boy did she fight back.

Disgraceful behaviour

After a protracted initial battle by Serena to get the video removed it kept getting reuploaded. Again and again she had to go through the same hoops to have it taken down.  Each time, the delays in removing the material were unconscionable but, as we  learn later, often Pornhub might remove the video but they left the tags on so anyone searching on the wider internet for that kind of thing would be brought back to their site. I’ll leave you to think about the “ethics” of that.

Serena reached out to Mike Bowe, a lawyer. He also appears before the Committee, giving detailed, broadly based forensic evidence which he put together, documenting years of deceitful if not illegal conduct by MindGeek.

Bowe’s evidence starts at 12:40. It is utterly astonishing and depressing. Why depressing? Because we also learned later, or were reminded, by the company themselves, they have been in existence since 2007 yet it is only since the article appeared in the New York Times a couple of months ago that they appear, or at any rate say, they are getting serious about dealing with illegal content.

Why anyone would believe anything these guys say is another matter and that applies equally to any so-called “transparency reports” they claim they are going to produce.

If ever there was a case for a strong, independent, legally-based public interest regulator to oversee online companies’ affairs it was made out in the Canadian Parliament this week. Well done to the MPs for keeping at it and well done to the Canadian Centre for Child Protection and others who never dropped the baton.

MindGeek speak for themselves

Mindgeek executives appeared before the Committee yesterday. I note in passing they say they employ 1,800 people (1,000 in Canada) and that 50% of their income is derived from ads. They frequently mention how many people, especially how many Canadians, visit their site as if this somehow gets them off whatever hook the Honourable Members of the Committee might think they are on. Of course it doesn’t. Or maybe MindGeek were threatening the MPs?

Dudes. We are very popular with your voters . Back off.”

Here is what appears in a newspaper to be a rather full account of what transpired when Mindgeek appeared at the Committee but, please  go back to the earlier link, tune in and listen. In particular listen to the MPs’  interrogation. Mr Erskine-Smith absolutely skewered them by asking  (around 13:20) one of the simplest and most obvious questions:

“How many times in 2020, 2019, 18 or 17 did individuals reach out to you to ask for a video to be taken down because it had been uploaded without their consent?”

Neither of the MindGeek executives who appeared before the Committee said they knew the number “off the top of their heads”. They could not or would not confirm that Serena Fleites had ever approached them.

They went to a public hearing of the Canadian Parliament and showed complete contempt for its proceedings.

The MindGeek executives made several references to how many social media platforms faced problems with child sex abuse content being uploaded.  That is true, they do, but none of the ones mentioned are established with the sole purpose of providing explicit sexual content which inevitably takes you to the edges of legality.

All social media companies have an ethical obligation to guard against illegal content, but if you are a porn based social media company it seems to me the obligation on you is that much higher. It is an obligation MindGeek did not meet.

My scribblings

I had submitted a brief note to the Canadian Committee. In it I focused not on the ethics of porn itself but on the ethics of setting up in and continuing as an online porn business. In the great scheme of things not particularly important, I suppose, but I reproduce it below in case any of you find any of its arguments useful in your ongoing battles.

“To many people the very idea of discussing whether or not a porn site can operate within an ethical framework will seem absurd because the porn industry itself is founded on unethical premises. I do not discuss that point. I limit myself to discussing the behaviour of porn companies, from an ethical standpoint.

It is well established that not everything that is legal is also necessarily ethical. In the case of  PornHub, and all commercial online porn companies of which I have any knowledge, what we typically see are investors realising there was a gap or an ambiguity in the law and in public policy then deciding to exploit it.

Many of the investors or their advisers were already in the porn business but all that does is underline the knowing nature of their ethical transgression. This is because,  and here my comments apply to amateur and commercial sites alike, they all knew or ought to have known that if they set up as porn providers in the physical world they would have been caught by and would have had to comply with long established and (relatively) effective, enforceable rules.

Thus,  by choosing to operate over the internet,  they were intentionally or recklessly ignoring and by-passing those limitations and norms. In effect they were smirking and saying “Catch me if you can”.

Cynically latching on to the new libertarian spirit of the internet age they, equally cynically, often donned the cape of free speech and artistic expression when really, all along, at least for the larger commercial players such as Mindgeek who are the main concern of your Committee,  it was just about making money. Alternatively porn publishers might claim, or it was claimed on their behalf, they had a valuable role to play in providing sex education.  It would be difficult to conjure up a more grotesque proposition.

The policy and legal gap or ambiguity which porn companies exploited emerged solely because the speed of technological change had completely outpaced the capacity of public policy makers and law makers to keep up. Some of the more far thinking porn merchants likely calculated that, eventually, public policy and the law would draw level but they would make a lot of hay while the sun continued to shine.

So while, in most jurisdictions,  porn providers may not have behaved illegally, they most certainly behaved unethically.

The reasons which lie behind the previously established  real world rules about access to porn did not vanish, nor were they reduced or materially altered just because the mode of delivery changed. On the contrary, the way the internet massively increased indiscriminate and unlimited access rather added to the ethical burden, a burden porn companies failed to discharge.

In the UK we had an analogous child protection problem in the late 1990s and early 2000s when  children started using online gambling web sites.  The legal age limit for gambling was and is 18. The same as it is to buy porn or go to a public cinema showing porn. Instances of a child being able to place a bet at a racecourse, at a football match or in a bookie’s shop were extremely rare for the simple reason the child could be seen and proof of age demanded. Penalties for failure to comply were severe.

When the internet arrived and gambling companies set up in cyberspace, every one of them acknowledged they were aware that children were using their services, placing bets via debit cards banks issued to account holders aged 12 or above. The gambling companies all said they were “very concerned” about the problem but actually almost all of them did nothing until the law compelled them to introduce age verification.

Once they introduced age verification we never heard of another case of a child simply ticking a box to say they were an adult and proceeding to gamble. The fact that a handful of gambling sites did take some steps to limit children’s access e.g. by disallowing debit cards they knew could be used by children,  rather amplified the ethical shortcomings  of the majority, who did nothing, claiming that asking everyone to tick the box showed they were doing their best.

The law requiring age verification to be introduced on gambling sites was passed in 2005 and became operative on 1st September 2007. Since that moment no company, or indeed any business providing “adult content” or age restricted goods such as  alcohol, tobacco and the like, had an ethical leg to stand on, at least not in any of the many countries where data sources exist which are similar to those in the UK.

Where such data sources do not exist porn companies and others could have invested in creating them as a prior condition of establishing or continuing to do business. Alternatively they could have ceased trading until they had developed an ethically sound system for keeping children away from their sites, and for preventing  adults from accidentally landing on their home page.  They did not do that. Like the UK’s gambling companies, the world’s porn companies are waiting to be forced to improve their behaviour. I hope Canada succeeds in bringing that about.”





Posted in Child abuse images, Consent, E-commerce, Privacy, Regulation, Self-regulation | Leave a comment

Regrets? I’ve had a few….

Last week (20th January) the UK Parliament’s Home Affairs Select Committee interviewed representatives of Facebook, WhatsApp, Twitter, Google, Snap and Tik Tok.

The Chair of the Home Affairs Select Committee is Yvette Cooper, an intellectual heavyweight of the first water. You had to feel a modicum of sympathy for the hapless folk the companies fielded. But only a modicum. A mini modicum.

Inevitably, on Inauguration Day in the USA, much of the Committee’s focus was on Trump, Trumpism and the post-truth world that helped create and sustain both. 6th January figured large.

To their credit none of the company representatives sought to deny or minimise the role  social media businesses played leading up to and including 6th January. The air was full of regrets for not acting sooner or differently. Phrases like “we are still learning”, “we know we must do better”, peppered the replies to MPs’ questions. All this put me in mind of Professor Sonia Livingstone’s aside to me in correspondence about the importance of

“breaking the cycle of

  1. Putting a new product or service into the  market
  2. Waiting for civil society to spot the problems and families to experience them
  3. Taking belated action.”

I might have added

4. Then being ready with self-deprecating comments like “we know we must do better” and “we are still learning“.

The disarming humility and contrition doubtless are genuinely meant at the time by the people speaking for their employers but humility and contrition butter no parsnips. Particularly when similar things keep on keeping on. There is a limit to the price societies can be expected to pay to allow companies the “freedom to innovate” . We are about to find out where that boundary lies. s230 is heading for the exit.

Facebook and end-to-end encryption

Yvette Cooper and others also raised questions about Facebook’s plans to introduce end-to-end encryption (E2E). In particular Cooper wanted to know what impact Facebook themselves  thought this would have on their own ability to detect child sex abuse images currently being exchanged via Messenger and Instagram Direct.

Monica Bickert’s reply was certainly truthful, in a literal sense, but it was also incomplete to the point of being deceptive. Her answer to Cooper’s question was

“I don’t know but I accept the numbers will go down”

Future hypotheticals

Bickert added that she thought the numbers would probably go down anyway because of other measures the company was taking. In other words the drop in numbers that is coming if things go ahead as planned may partly be down to Facebook simply being more effective in discouraging illegal behaviour which threatened or harmed children. Cooper exposed this as self-exculpating baloney.

Turns out it largely hinges or depends on planned educational initiatives designed to help children avoid abusive individuals and situations in the first place.  Not exactly mind-blowing or revolutionary. In fact it is the kind of stuff they are already doing and if all Bickert is saying is they will do more of it or better then bring it on. It is welcome even though a tad oblique as compared with straightforward detection, deletion and reporting, which was the main thrust of Cooper’s questioning. Cooper was not asking about images that might not be created or exchanged or paedophiles who might be avoided.

46% decline in 21 days

Cooper referred to numbers published some time ago by NCMEC. These suggested if Facebook went ahead with E2E there could be a 70% drop in images being detected, deleted and reported. That’s globally.

What Cooper evidently did not know, but Bickert  must have, was the day before the Select Committee meeting NCMEC had published new data showing the known or actual effect of Facebook ceasing to be able to detect child sex abuse in the manner they had hitherto.

Because of the fiasco with the European Electronic Communications Code, on 20th December in all EU Member States Facebook stopped scanning  for child sex abuse materials. Stopping scanning has exactly the same effect as introducing E2E.

On 19th January, NCMEC’s new published numbers showed in the 21 days immediately following 20th December there had been a 46% drop in reports from EU countries.

Excluding the UK, in the three weeks prior to 20th December NCMEC received 24,205 reports  linked to EU Member States. In the three weeks afterwards it dropped to 13,107. We will never know which children were in the 11,000 images that weren’t  picked up.  How many were new images, never seen before, with all that that entails?

So when Cooper asked, as she did twice, about the likely effect of introducing end-to-end encryption Bickert was truthful when she said she couldn’t say but she might have at least mentioned the numbers NCMEC had just published. Then she could have explained why a 46% drop, or worse, concretely, not hypothetically, is a price worth paying.

Facebook blames their customers

Cooper persistently challenged Bickert as to why they were going ahead with E2E at all when they knew it will mean more children will be put in harm’s way, more perpetrators will go un-caught and un-punished. Bickert’s answer was, er, “surprising”. 

Bickert referred to a survey of British adults who, seemingly, listed privacy related concerns as their “top three”. I am not sure which survey Bickert had in mind, she didn’t say, but if it was the 2018  Ofcom one she might have read a little further and seen “the leading area of concern” is the protection of children. But even if that was not the case,  whether or not children were “listed” in the top 50 concerns expressed by adults, teens or stamp collectors for that matter, what was  Bickert really saying?

“Don’t blame us. We’re only doing this because it’s what the dudes want and our job is to give it to them.”

An industry standard?

Bickert and her colleague from WhatsApp shifted their ground a little saying “strong encryption is now the industry standard” as if this was the key justification for going ahead with or retaining E2E.  Cooper pointed out that Facebook was a major part of the industry so that amounted to rather transparent, self-serving circular reasoning. Moreover in other areas Facebook has  repeatedly shown it is willing to strike out alone and not just follow the herd. They cannot now shelter behind the actions of others.

The underlying reasons?

Suggesting  something is an “industry standard” is simply a less vulgar or less pointed way of saying “our revenues will likely be badly impacted if we don’t do this”. It’s a variation on the dudes theory expounded earlier. In other words it is about money.

Secondly, how did we get to a point where the dudes seemingly feel they need to have E2E? Isn’t it because of the previous actions and admitted failures of companies like Facebook?

So first they create the problem and then they come up with the wrong answer to it. Chutzpah on stilts.

Facebook’s “pivot to privacy” is alliteratively admirable but not in any other way. It is about Facebook trying to repair its appalling image in the privacy department, based on its history of not respecting or taking sufficient care of its users’ privacy. It is acting now in order to continue generating gigantic quantities of moolah.

Towards a very dark place

We may never know what role encrypted messaging services played in organizing and orchestrating the events of 6th January but few can doubt that the unchecked growth of strongly encrypted messaging services is taking us towards a very dark place. A place where child abusers as well as fascist insurrectionists feel safe.

In and of itself strong encryption is not a bad thing. Indeed it is now essential in many areas.  But in the wrong hands, used for the wrong purposes, it can facilitate a great deal of serious damage. We have to find a way to ensure that does not happen. If companies like Facebook do not find a way of doing that, they will have one thrust upon them. The Silicon Valley experiment has run its course. It will soon look different.

Posted in Child abuse images, Facebook, Google, Internet governance, Regulation, Self-regulation | Leave a comment

Adding insult to irony

If, like me, you were brought up a Catholic or in another Christian denomination, likely you will know 6th January is celebrated by the faithful as the “Feast of the Epiphany. Well, in a secular sense, 6th January 2021 definitely was an epiphany for us all, meaning a moment of profound revelation.

Witness five dead bodies in or around the US Congress and a televised attempt to frustrate the outcome of an election in order to preserve in office a liar and a cheat who openly incited violence while actively seeking to undermine the Constitution.

Of course the events of that day did not come out of nowhere. They reflect a deeper malaise and divisions. Rooted in disillusionment, the American Dream is not delivering for them, a great many angry people found confirmation bias in the constant stream of falsehoods and distortions fed to them by Trump and his fellow conspirators.  Never have the consequences of allowing a “Post-Truth” society to emerge and grow been more clearly in evidence.

Can there be any real doubt about the role social media companies played in creating, sustaining and amplifying the societal fissures that brought us to 6th January? Let’s not get into the practical, organizing role social media also played in orchestrating the murderous assault. That’s for another day. Will we ever know how much was done through strongly encrypted channels? Probably not.

It doesn’t stop there

The aftermath of 6th January 2021 then saw private entities, companies, silencing the President of the United States and effectively shutting down a speech app (Parler) altogether or very substantially.  In so doing Silicon Valley added insult to irony.

They gave Trumpism a megaphone and, in the name of free speech, timorously stood back, letting it blossom as the dollars rolled in. It was only when Trump went almost foaming-at-the-mouth insane and the scenes of 6th January were televised, that the inescapable and repeatable logic of the laissez faire s.230 nightmare was fully, unavoidably exposed.

Then some of the same companies decided to shut Trump up. It’s hard to think of this either as a step too far or as a step in the right direction because in a better and more rational world the need for it to be taken at all would never have arisen. What started as a benign experiment with technology brought the USA, and therefore the world, to the edge of disaster.

The amount of sympathy I have for Trump or Parler can be measured only in large minus quantities. That is not the point. What is the point is the egregious presumption of private bodies deciding to make public policies in areas of fundamental importance to our whole way of life.  De haute en bas they float above us mere mortals and tell us when we meet their  exacting standards and punish us when we don’t. Sadly they are constantly deflected by the desire to earn money. This clouds their vision from time to time.

Fundamentally this is a failure of governance

A great many idealistic people who were disgusted with the shortcomings of mainstream politics either in their own country or globally, or both, saw the internet as a way of establishing a whole new set of possibilities.

Then the money moved in. The money saw different opportunities and did something really smart. Cynical, but smart. Not only did they get s. 230 adopted in the USA and copied elsewhere, they also managed to implant in people’s minds the idea that the absence of regulation was the same as “freedom”.  Any attempt to regulate the  internet  (meaning them or their businesses) was portrayed as an actual or potential attack on  “freedom”.  Politicians  and judges stepped back. Unsure of themselves. In truth the absence of regulation was just another way of creating room to make more cash.

After the money came the totalitarians. They learned a lot from what they observed elsewhere. In particular they learned from surveillance capitalism. Often the very same companies and engineers that helped Palo Alto were now helping Pyonyang.

Meanwhile we have a UN body called the Internet Governance Forum which, since 2006,  has pretended to have some influence on matters of the kind discussed here. I predict it is not long for this world.  It has been coming for a while. 6th January sealed its fate. That’s a shame in many ways because the Forum has great strongpoints.

Mozilla’s plans to encrypt DNS queries in Firefox 

What has the main argument I am making in this blog got to do with child protection? Everything. If you doubt that just read a consultation document published by Mozilla. In particular look at this sentence:

“Numerous ISPs today provide opt-in filtering control services, and  (we intend) to respect those controls where users have opted into them.”  (emphasis added).

To put that slightly differently,  Mozilla has decided not to “respect” those controls where users have not “opted into them”.

A self-appointed techno-priesthood  has decreed that one approach to child protection is acceptable and another is not. Can I resist pointing out Mozilla’s global HQ is in a place called “Mountain View”? No I cannot. I do so as a service for those wondering where the latterday Olympus is to be found.

Inertia is at the root of many evils in the internet space, particularly among the less literate and  less knowledgeable, people who are often also among the most vulnerable e.g. children. Whatever an individual ISP may have lawfully decided to do, Mozilla seem to be willing to expose children to the risk of harm unless and until their parents get their act together and choose to opt in to protective filters. Wrong answer.  By a mile.

Mozilla’s  consultation document was written before 6th January, 2021. What it truly shows is Zeus needs to go back to the drawing board.

Posted in Internet governance, Regulation, Self-regulation | Leave a comment

Absurdities and atrocities

Voltaire famously said “Those who can make you believe absurdities, can make you commit atrocities.” History is littered with examples of this and last week in the USA we saw the same thing played out again.

Lies and travelling trousers

In the age of the internet never was it more true that a lie can be half way around the world before the truth has got its trousers on. The more fanciful or ridiculous (absurd) the lie the faster it is likely to spread through social media platforms. Eyeballs mean money and money is the name of their game.

We need urgently to get over the initial, marvellous hippy notion that in the internet we created something that enables everyone to be a publisher, a journalist, a doughty warrior concerned only to make the world a better place. That is true, we have.

But it is now abundantly clear we have also created something  which threatens that very idea. Last week was the proof, played out on TV.

How far are we willing to go to defend the world that emerged from and through the post-War settlement? The fate of Weimar should not be forgotten or what followed.

Oh the irony!

I am not the first person to note or comment on the irony. Governments have threatened to regulate social media platforms but now we see social media platforms doing something that looks very like regulating Governments.

Of course in a narrow way you could argue depriving Trump of his Twitter account or banning him from Facebook until he ceases to be President is not directly regulating a Government as such, but it is so close you would be hard-pressed to insert a Rizzla paper between the two.

Too little too late

Obviously, I approve of what Twitter and Facebook did but that isn’t the point. One might ask why they didn’t do it a lot sooner. But the larger questions are how it ever came to this in the first place and could it happen again?

Trump and his cronies incited the mob in an assault on democracy, but he and they could only get to a point where that was possible because social media platforms and elements of the mainstream media helped build him up. The USA is now on a national alert because of fears similar acts will be repeated in State Capitols on 20th January. Inauguration Day.

The intimacy, immediacy and scale of the internet made a “post-truth” society possible. We have had lying politicians and lying campaigners before, but  in modern times we have never had lying politicians or campaigners who had the financial backing and tools such as the internet, with its handmaiden, profiling, to enable dangerous demagogues to reach and manipulate, the lumpen, the alienated-dispossessed, the angry and the frightened.

Preserving liberal values is about preserving decency

So, yes, of course, we should have serious discussions about what free speech means in the age of the internet but if liberal democracy and liberal values are threatened where does it say we must stand by and let them die because we are paralyzed by anxiety and by laws drawn up for entirely different times? We need a legal framework which comprehends and embraces life in the early 21st Century. And we need it sooner, not later.

Posted in Facebook, Google, Internet governance, Privacy, Regulation, Self-regulation | Leave a comment

We need to treat Treaties seriously

The North American Free Trade Association (NAFTA) was a free trade agreement between the USA, Canada and Mexico. It became operative in 1994, although free trade between two of the three, Canada and the USA, had existed since 1989.

In 2020 NAFTA was updated by the US-Mexico-Canada Agreement (USMCA). One of the principal differences between NAFTA and USMCA was the inclusion in USMCA of a raft of provisions which set limits on what any of the three countries can do in respect of policies impacting on the internet. In 1989 and 1994 the internet was still in its infancy, a long way from the massive and pervasive presence it has today.

USMCA is not an agreement made by three parties of equal status and power

Lest anyone runs a way with the idea the USA felt obliged to agree to conditions advanced by the Canadians or the Mexicans following discussions between equals, even the most cursory glance at the terms of the deal makes it clear the USA was calling all the key shots. This is not surprising given the size and value of their economy.

Here are some explanatory extracts from the article linked to above

“Unfortunately, not everyone is free to experience the full benefits of the internet age, because so many national governments restrict access to online services and websites…. (emphasis added).

To counter this trend, the USMCA prohibits a broad range of digital trade restrictions…. (ditto)

The agreement’s most important digital trade provisions enshrine policies essential to the effectiveness and operation of the global internet….

Another essential policy mandated by the USMCA is liability protection for online intermediaries. It may seem arcane and technical, but liability protection is a core policy of the laissez-faire approach that enabled the digital revolution to occur in the first place. In the United States, laws like the Digital Millennium Copyright Act and Section 230 of the Communications Decency Act prevent website operators from being held liable for the conduct of users on their sites….. (ditto)

……. liability protections are a key reason why the world’s most successful and innovative internet companies were built in America. (ditto)

The authors of the article did not add, but I will, liability protection is a major reason why the internet is the way it is today. Think “move fast and break things.”

Digital businesses were given unique privileges.  At the time the rules were being set the technology itself and the likely dynamics of the way the internet would develop were poorly understood by policy makers and judges. Thus companies were not incentivised to root out and prevent harm. On the contrary they were incentivised to wrap themselves up and remain inert in the warm blanket of immunity.

Are you getting the picture? So to speak

How did all these clauses and provisions get included in USMCA? Big Tech lobbied for them.  By that I mean Silicon Valley. American Big Tech. They wanted to export the American way, their way.

A more naked statement of self-interest it would be hard to find. Never mind what ideas these pesky “national governments” might get into their heads. “It’s our way or the highway” was the message. US companies were particularly concerned to get certain clauses in USMCA because the Canadian Government and the Canadian Supreme Court in particular were starting to get their act together and be more assertive. They needed putting back in their box.

Nice work if you can get it

Do I blame American tech companies for seeking to further their own interests? No. I do not. Why wouldn’t they? I might have a slight complaint that very often the lobbying done on their behalf was  executed by or through well-funded intermediaries like trade associations because the individual firms did not want to become identified in the public eye with such stark expressions of commercial self-interest and/or nationalist sentiment while, because of a shortage of resources, children’s organizations were not able to mount an effective counter lobby. But hey! Whoever said life was fair? The playing field is not level across a vast acreage, not just here.

So this Wednesday in the UK Parliament

Following the publication last month of the UK Government’s  final response to the consultation on online harms, the UK is about to embark upon its own online regulatory odyssey. I will be writing about that odyssey in due course, meaning soon, but let me say now there is much in what the Government is proposing which I heartily welcome. And a great deal of it would be completely nullified or circumscribed if US interests managed to do a repeat performance in respect of the UK-US trade negotiations currently underway. Our world-leading Age Appropriate Design Code would also be threatened.

We must not let that happen which is why, in the House of Lords on Wednesday 6th January 2021, it is very much to be hoped that as many Peers as possible  will vote in favour of the cross-party amendment standing in the name of Baroness Kidron, Lord Stevenson of Balmacara, Lord Clement-Jones and Lord Sheikh. A big majority in the Lords will encourage  Members in the Commons to get behind it when it reaches them.

PS Just in case it isn’t obvious, similar considerations would apply to any trading bloc or individual country that enters into discussions with the USA about a free trade agreement.

Posted in Age verification, Default settings, Internet governance, Regulation, Self-regulation | Leave a comment

A very bad day for children in Europe

If you live in an EU Member State and you have used Facebook Messenger or Instagram Direct today you probably saw this message. “Some features are not available. This is to respect new rules for messaging services in Europe. We’re working to bring them back.”

This cryptic statement refers, among other things, to the fact Facebook have turned off their proactive child protection tools in every EU Member State.  This is because today the provisions of the European Electronic Communications Code (EECC) kick in.

But Microsoft, Google, Linkedin, Roblox and Yubo somehow managed to find a way to carry on with the tools. Well done them.

Given Facebook is the world’s largest source of csam and other child sexual exploitation materials reported to NCMEC and law enforcement, this is unbelievably disappointing.

This should never have happened in the first place BUT

OK, we should never have got into this position but where there is a will there is a way. Obviously with the five companies I just named there was a will to carry on protecting children by continuing to use the tools. They did find areas of doubt sufficient to justify a continuation. Facebook didn’t.

Facebook is usually not slow to act when an important commercial interest is threatened. Not here. Facebook rolled over.

Facebook is trying to reshape its image

Facebook is determined to appease and reach out to the privacy lobby. That is plainly an overriding corporate objective that trumps all others. Given the company’s previous lack of care and respect for their users’ privacy it is not hard to work out why they want to reposition themselves  in this way.

But children are paying the price for their inglorious corporate history.

Until this is put right – as it surely will be – how many pictures of children being sexually abused will continue to circulate on the internet? How many paedophiles will manage to connect with children?  How many existing victims will be harmed further, and how many new victims will there be? We will never know, but is unlikely to be zero.

Does Facebook really still have a Safety Advisory Board? Were they consulted about this, if so when and what did they say?

The anti-suicide and self-harm tools?

What about the tools which try to detect a child contemplating suicide or self-harm? Have they also been suspended? Maybe they haven’t but essentially they work in the same way as the anti-grooming tools and the classifiers used to detect possible csam. Facebook should put out a statement specifically commenting on that point.

Concrete results

Last month NCMEC published a letter  to MEPs in which they gave some hard numbers.

In 2019  NCMEC received 16.9 million reports referencing 69 million items of csam or child sexual exploitation. Of these “over 3 million” originated in the EU. That is 4% of the total, or about 250,000 per month. 95% of these reports came from services affected by the  EECC.  From these reports 200 children living in Germany were identified, as were 70 children living in Holland. In the same letter we see the 2020 numbers are going to be higher.

Knowing Facebook accounts for the great majority of reports to which the NCMEC letter refers, we can see the likely dimensions of what Facebook have done.

Shame on Facebook. Let’s hope they succeed in “bringing them back” as soon as possible. Then they can announce they are dropping or modifying their plans to encrypt the very same services.

UK exempt?

Why do the tools continue in use in the UK?  It seems because we adopted laws at national level which provide a good enough legal basis. Can it really be the case that no other Member State did the same? And if one or more did how can Facebook justify cutting them off?

This has been a bad day for children in Europe.

We are heading for a strange world

Privacy laws were never intended to make it easier for paedophiles to connect with children. They were never intended to make it easier for pictures of children being raped to be stored or circulated online. And it would be a strange world indeed if that is where we are heading.

If there truly is a legal problem here it cannot be one of substance. It can only have arisen because various bureaucrats and lawyers did not get all their ducks in a row and take all the right steps at the right time.

Instead of a brave stance in defence of children, Facebook has buckled in front of the remediable incompetence of others.

Posted in Child abuse images, Default settings, Facebook, Google, Microsoft, Privacy, Regulation, Self-regulation | 1 Comment

A new industry award

On this crucial day for children, as the EU’s “trilogue” meets to decide the fate of proactive child protection tools within the 27 Member States, I have decided to inaugurate a new annual award.

I think I will call it “The Techno Chutzpah Oscar” but if anyone can come up with a better name please let me know.

The Oscar will go to the company that most transparently and egregiously behaves or speaks hypocritically in the context of online child protection. And I have no hesitation naming Facebook the inaugural winner.

Here is an extract from the New York Times of 4th December 2020

“Facebook, the most prolific reporter of child sexual abuse imagery worldwide, said it would stop proactive scanning entirely in the E.U. if the regulation took effect. In an email, Antigone Davis, Facebook’s global head of safety, said the company was “concerned that the new rules as written today would limit our ability to prevent, detect and respond to harm,” but said it was “committed to complying with the updated privacy laws.” ( emphasis added)

This statement came from the company that normally goes straight to court when it decides it doesn’t like something a Government has said or done. Such legal actions sometimes win, sometime lose, but they almost always delay. Why not here? Why the immediate collapse without so much as the whiff of a writ?

Instead of  Facebook saying it is

“committed to complying with the updated privacy laws” 

could we not have heard the following?

“We saw the Opinion of the EDPS and we think it is rubbish. Facebook  believes there is a clear and firm legal basis which supports our use of proactive child protection tools. Our lawyers wouldn’t have let us deploy them in the first place were it otherwise. This legal basis is established under a variety of international legal instruments. In fact we would go further and say we believe we have both a legal and a moral obligation to use the best available means to protect children. We will vigorously defend that position in court should it prove necessary.”

But maybe the more obvious point, the one that gets them over the line and justifies the award of the first ever Techno Chutzpah Oscar is, lest we forget, Facebook  is the company that has acknowledged it has gigantic quantities of child sex abuse imagery being exchanged using its platforms but, nevertheless, still intends to encrypt the very services the new EU privacy law affects, if it remains unaltered.

If Facebook goes ahead with end-to-end encrytion in the way they have said, what happens with the EU law will not matter, at least not within EU Member States, because none of the tools will be able to penetrate the encrytion anyway.

Am I being unkind and cyncial? Was Facebook merely striking a pose to try to encourage the EU to do the right thing, in part because they have already decided internally to abandon end to end encryption? Answers on a postcard please to the usual address.

Posted in Child abuse images, E-commerce, Privacy, Regulation, Self-regulation | Leave a comment

Half a pat on the back

Thanks to tremendous lobbying and campaigning work by children’s organizations from across the world we have won the first part of what we wanted to achieve.

LIBE says “yes”

MEPs were tremendously impressed by the breadth and scale of support there was for the positions we took up on the Commission’s proposed derogation. It strengthened the hands of our friends in the European Parliament and hugely weakened our opponents.

The LIBE Committee today voted to put forward a report to the plenary meeting of the European Parliament next week. That means it should be possible for the Trialogue to meet and decide the matter in time to beat the 20th December deadline.

Here is the press release of the Child Rights Intergroup welcoming the decision.

But we can only give ourselves half a pat. There is more still to be done.

I say this because from the press release issued by the Parliament and from other reports, there are bits of what the Committee appear to have agreed which could still derail us.

No interruption or suspension of the tools

Conditions or riders have been attached to the continued use of the tools. If that means the tools are suspended for any period of time these conditions or riders must be resisted.

The conditions and riders are about transparency, accountability and reporting. These are things children’s groups should be very strongly in favour of but, at this late stage, to say they must be sorted out as a condition of continuing to use the tools seems utterly wrong.

So my suggestion is, over the next few days and into next week, we continue to lobby MEPs and national Governments – particularly the German Government – saying something along these lines:

  1. It is vital the Trialogue completes its work ahead of the 20th December deadline.
  2. We are concerned, however, that, even if the Trialogue does complete its work in time, if the LIBE decision is followed in total the use of the tools may be made conditional on terms that almost certainly cannot be met within such a short timescale.
  3. We have no problem or objection to stipulations about accountability, transparency or reporting mechanisms attaching to the continued use of the tools by companies. On the contrary we welcome them, but the only reasonable course of action is to allow these matters to be resolved during the period of grace which the derogation will establish or as part of the longer term strategy if that is adopted during the period of grace.

We can then turn our attention to what happens during the period grace and, above all, we can start to focus on working out what a long term policy will look like.

Here is NCMEC’s statement which also discusses that point.

Posted in Child abuse images, Privacy, Regulation, Self-regulation | 2 Comments

The questions to be asked in Brussels

Crunch time approaches in Brussels. Members of the LIBE Committee and later the plenary need to focus on the following questions:

  1. When the GDPR was making its way through the European Instistitions do you think the co-legislators expressly intended to make it impossible for tech companies to prevent their customers from publishing, exchanging or storing images (still pictures or videos) of children being raped?
  2. When the GDPR was making its way through the European Instistitions do you think the co-legislators expressly intended to prevent or delay the identification and removal from public view of images of children being raped?
  3. When the GDPR was making its way through the European Instistitions do you think the co-legislators expressly intended to make it easy for sexual predators to locate and engage with children?
  4. When the GDPR was making its way through the European Instistitions do you think the co-legislators expressly intended to prevent companies from trying to identify children who might be contemplating suicide or self-harm so as to divert them from that path?

I believe the answer to all of these questions is a simple, unqualified “no”.

Are there ways of deploying the kinds of child protection tools referred to which are entirely and unequivocally compliant with the highest privacy standards?

I believe the answer to that question is a simple, unqualified “yes”.

So now I am an MEP

Let’s say I am a Member of the LIBE Committee,  from Poland or Ireland – I am  an Irish citizen and I could become a Polish citizen. I am 100% in favour of protecting children to the greatest extent possible.  But what do I see?

A lack of transparency and safeguards

I have no evidence any company has behaved inappropriately or put anyone in danger, child or adult, when processsing data that might be associated with the deployment of child protection tools. All I have is a deeply rooted suspicion. Call it a hunch.

This deeply rooted suspicion was allowed to take hold and flourish because there is no trusted transparency regime with associated safeguards and metrics emanating from accountable public sources which could assure me all is well.

I am asked to take everything on trust. That is wrong. No other word for it and it must be addressed in the forthcoming Digital Services Act. But what do I do in the meantime?

Two wrongs do not make a right

I look at how poorly some individual Member States have responded to the child protection challenge, as evidenced, for example, by their failure to implement fully the terms of the 2011 EU Directive  but also by their failure to act more broadly in society at large where the bulk of child sex abuse and threats to children occur.

I conclude the national politicians responsible, maybe even in my own Party, are only paying lip service to the idea of protecting children.

I look at the patchy engagement of some law enforcement agencies.

I look at how the different child protection tools we are discussing have emerged from private tech companies, starting back in 2009, and finally I look at what I think are the failures of Commission officials and Member States to address all these things satisfactorily up to now.

I might even reflect on my own responsibility here. This is not my first term as an MEP.

But still. What do I do?

Having looked at what I  believe is a series of process failures and other shortcomings do I then decide my higher duty is to those processes? Do I vote to bring an end to the tools?  Even for a short while until the mess is sorted out and all the procedural ducks are in a neat, bureaucratically satisfying row?  Should I vote to throw out the Commission’s interim proposal? Should I refuse to give children the benefit of the doubt?

Absolutely not

Posted in Child abuse images, Internet governance, Privacy, Regulation, Self-regulation | Leave a comment