net.wars

"What's that for?" I asked. The question referred to a large screen in front of me, with my newly-captured photograph in the bottom corner. Where was the camera? In the picture, I was trying to spot it.

The British Airways gate attendant at Chicago's O'Hare airport tapped the screen and a big green checkmark appeared.

"Customs." That was all the explanation she offered. It had all happened so fast there was no opportunity to object.

Behind me was an unforgiving line of people waiting to board. Was this a good time to stop to ask:

- What is the specific purpose of collecting my image?

- What legal basis do you have for collecting it?

- Who will be storing the data?

- How long will they keep it?

- Who will they share it with?

- Who is the vendor that makes this system and what are its capabilities?

It was not.

I boarded, tamely, rather than argue with a gate attendant who certainly didn't make the decision to install the system and was unlikely to know much about its details. Plus, we were in the US, where the principles of the data protection law don't really apply - and even if they did, they wouldn't apply at the border - even, it appears, in Illinois, the only US state to have a biometric privacy law.

I *did* know that US Customs and Border Patrol had begun trialing facial recognition in selected airports, beginning in 2017. Long-time readers may remember a net.wars report from the 2013 Biometrics Conference about the MAGICAL [sic] airport, circa 2020, through which passengers flow unimpeded because their face unlocks all. Unless, of course, they're "bad people" who need to be kept out.

I think I even knew - because of Edward Hasbrouck's indefatagable reporting on travel privacy - that at various airports airlines are experimenting with biometric boarding. This process does away entirely with boarding cards; the airline captures biometrics at check-in and uses them to entirely automate the "boarding process" (a favorite bit of airline-speak of the late comedian George Carlin). The linked explanation claims this will be faster because you can have four! automated lanes instead of one human-operated lane. (Presumably then the four lanes merge into a giant pile-up in the single-lane jetway.)

It was nonetheless startling to be confronted with it in person - and with no warning. CBP proposed taking non-US citizens' images in 2020, when none of us were flying, and Hasbrouck wrote earlier this year about the system's use in Seattle. There was, he complained, no signage to explain the system despite the legal requirement to do so, and the airport's website incorrectly claimed that Congress mandated capturing biometrics to identify all arriving and departing international travelers.

According to Biometric Update, as of last February, 32 airports were using facial recognition on departure, and 199 airports were using facial recognition on arrival. In total, 48 million people had their biometrics taken and processed in this way in fiscal 2021. Since the program began in 2018, the number of alleged impostors caught: 46.

"Protecting our nation, one face at a time," CBP calls it.

On its website, British Airways says passengers always have the ability to opt out except where biometrics are required by law. As noted, it all happened too fast. I saw no indication on the ground that opting out was possible, even though notice is required under the Paperwork Reduction Act (1980).

As Hasbrouck says, though, travelers, especially international travelers and even more so international travelers outside their home countries, go through so many procedures at airports that they have little way to know which are required by law and which are optional, and arguing may get you grounded.

He also warns that the system I encountered is only the beginning. "There is an explicit intention worldwide that's already decided that this is the new normal, All new airports will be designed and built with facial recognition built into them for all airlines. It means that those who opt out will find it more and more difficult and more and more delaying."

Hasbrouck, who is probably the world's leading expert on travel privacy, sees this development as dangerous. Largely, he says, it's happening unopposed because the government's desire for increased surveillance serves the airlines' own desire to cut costs through automating their business processes - which include herding travelers onto planes.

"The integration of government and business is the under-noticed aspect of this. US airports are public entities but operate with the thinking of for-profit entities - state power merged with the profit motive. State *monopoly* power merged with the profit motive. Automation is the really problematic piece of this. Once the infrastructure is built it's hard for airline to decide to do the right thing." That would be the "right thing" in the sense of resisting the trend toward "pre-crime" prediction.

"The airline has an interest in implying to you that it's required by government because it pressures people into a business process automation that the airline wants to save them money and implicitly put the blame on the government for that," he says. "They don't want to say 'we're forcing you into this privacy-invasive surveillance technology'."

Illustrations: Edward Hasbrouck in 2017.

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Stories about the border wars between cyberspace and real life are posted occasionally during the week at the net.wars Pinboard - or follow on Twitter.

Posted by Wendy M. Grossman at 3:34 PM | Permalink | Comments (0) | TrackBacks (0)

In the spring of 2020, as country after country instituted lockdowns, mandated contact tracing, and banned foreign travelers, many, including Britain, hastily passed laws enabling the state to take such actions. Even in the strange airlessness of the time, it was obvious that someday there would have to be a reckoning and a reevaluation of all that new legislation. Emergency powers should not be allowed to outlive the emergency. I spent many of those months helping Privacy International track those new laws across the world.

Here in 2022, although Western countries believe the acute emergency phase of the pandemic is past, the reality is that covid is still killing thousands of people a week across the world, and there is no guarantee we're safe from new variants with vaccine escape. Nonetheless, the UK and US at least appear to accept this situation as if it were the same old "normal". Except: there's a European war, inflation, strikes, a cost of living crisis, energy shortages, and a load of workplace monitoring and other privacy invasions that would have been heavily resisted in previous times. (And, in the UK, a government that has lost its collective mind; as I type no one dares move the news cameras away from the doors of Number 10 Downing Street in case the lettuce wins.)

Laws last longer than pandemics, as the human rights lawyer Adam Wagner writes in his new book, Emergency State: How We Lost Our Freedoms in the Pandemic and Why It Matters. For the last couple of years, Wagner has been a constant presence in my Twitter feed, alongside numerous scientists and health experts posting and examining the latest new research. Wagner studies a different pathology: the gaps between what the laws actually said and what was merely guidance. and between overactive police enforcement and people's reasonable beliefs of what the laws should be.

In Emergency State, Wagner begins by outlining six characteristics of the power of emergency-empowered state: mighty, concentrated, ignorant, corrupt, self-reinforcing, and, crucially, we want it to happen. As a comparison, Wagner notes the surveillance laws and technologies rapidly adopted after 9/11. Much of the rest of the book investigates a seventh characteristic: these emergency-expanded states are hard to reverse. In an example that's frequently come up here, see Britain's World War II ID card, which took until 1952 to remove, and even then it took Harry Wilcock to win in court after refusing to show his papers on demand.

Most of us remember the shock and sudden silence of the first lockdown. Wagner remembers something most of us either didn't know or forgot: when Boris Johnson announced the lockdown and listed the few exceptional circumstances under which we were allowed to leave home, there was as yet no law in place on which law enforcement could rely. That only came days later. The emergency to justify this was genuine: dying people were filling NHS hospital beds. And yet: the government response overturned the basis of Britain's laws, which traditionally presume that everything is permitted unless it's specifically forbidden. Suddenly, the opposite - everything is forbidden unless explicitly permitted - was the foundation of daily life. And it happened with no debate.

Wagner then works methodically through Britain's Emergency State, beginning by noting that the ethos of Boris Johnson's government, continuing the conservatives' direction of travel, coincidentally was already disdainful of Parliamentary scrutiny (see also: prorogation of Parliament) and ready to weaken both the human rights act and the judiciary. As the pandemic wore on, Parliamentary attention to successive waves of incoming laws did not improve; sometimes, the laws had already changed by the time they reached the chamber. In two years, Parliament failed to amend any of them. Meanwhile, Wagner notes, behind closed doors government members ignored the laws they made.

The press dubbed March 18, 2022 Freedom Day, to signify the withdrawal of all restrictions. And yet: if scientists' worst fears come true, we may need them again. Many covid interventions - masks, ventilation, social distancing, contact tracing - are centuries old, because they work. The novelty here was the comprehensive lockdowns and widespread business closures, which Wagner suggests may have come about because the first country to suffer and therefore to react was China, where this approach was more acceptable to its authoritarian government. Would things have gone differently had the virus surfaced in a democratic country? We will never know. Either way, the effects of the cruelest restrictions - the separation among families and friends, the isolation imposed on the elderly and dying - cannot be undone.

In Britain's case, Wagner points to flaws in the Public Health Act (1984) that made it too easy for a months-old prime minister with a distaste for formalities to bypass democratic scrutiny. He suggests four remedies: urgently amend the act to include safeguards; review all prosecutions and fines under the various covid laws; codify stronger human rights, either in a written constitution or a bill of rights; and place human rights at the heart of emergency decision making. I'd add: elect leaders who will transparently explain which scientific advice they have and haven't followed and why, and who will plan ahead. The Emergency State may be in abeyance, but current UK legislation in progress seeks to undermine our rights regardless.

Illustrations: The Daily Star's QE2 lettuce declaring victory as 44-day prime minister Liz Truss resigns.

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Stories about the border wars between cyberspace and real life are posted occasionally during the week at the net.wars Pinboard - or follow on Twitter.

Posted by Wendy M. Grossman at 7:28 PM | Permalink | Comments (0) | TrackBacks (0)

A while back, I was trying to get a friend to install the encrypted messaging app Signal.

"Oh, I don't want another messaging app."

Well, I said, it's not *another* messaging app. Use it to replace the app you currently use for texting (SMS) and it will just sit there showing you your text messages. But whenever you encounter another Signal user those messages will be encrypted. People sometimes accepted this; more often, they wanted to know why I couldn't just use WhatsApp, like their school group, tennis club, other friends... (Well, see, it may be encrypted, but it's still owned by the Facebook currently known as Meta.)

This week I learned that soon I won't be able to make this argument any more, because...Signal will be dropping SMS support for Android users sometime in the next few months. I don't love either the plan or the vagueness of its timing. (For reasons I don't entirely understand, this doesn't apply to the nether world of iPhone users.)

The company's blog posting lists several reasons. Apparently the app's SMS integration is confusing to many users, who are unclear about when their messages are encrypted and when they're not. Whether this is true is being disputed in the related forum thread discussing this decision. On the bah! side is "even my grandmother can use it" (snarl) and on the other the valid evidence of the many questions users have posted about this over the years in the support forums. Maybe solvable with some user interface tweaks?

Second, the pricing differential between texting and Signal messages, which transit the Internet as data, has reversed since Signal began. Where data plans used to be rare and expensive, and SMS texts cheap or bundled with phone service, today data plans are common, and SMS has become expensive in some parts of the world. There, the confusion between SMS and Signal messaging really matters. I can't argue with that except to note that equally it's a problem that does *not* apply in many countries. Again, perhaps solvable with user settings...but it's fair enough to say that supporting this may not be the best use of Signal's limited resources. I don't have insight into the distribution of Signal's global user base, and users in other countries are likely to be facing bigger risks than I am.

Third is sort of a purity argument: it's inherently contradictory to include an insecure protocol in an app intended to protect security and privacy. "Inconsistent with our values." The forum discussion is split on this. While many agree with this position, many of the rest of us live in a world that includes lots of people who do not use, and do not want to use (see above), Signal, and it is vastly more convenient to have a single messaging app that handles both.

Signal may not like to stress this aspect, but one problem with trusting an encrypted messaging app in the first place is that the privacy and security are only as good as your correspondents' intentions. Maybe all your contacts set their messages to disappear after a week, password-protect and encrypt their message database, and assign every contact an alias. Or, maybe they don't password-protect anything, never delete anything, and mirror the device to three other computers, all of which they leave lying around in public. You cannot know for sure. So a certain level of insecurity is baked into the most secure installations no matter what you do. I don't see SMS as the biggest problem here.

I think this decision is going to pose real, practical problems for Signal in terms of retaining and growing its user base; it surely does not want the app's presence on a phone become governments' watch-this-person flag. At least in Western countries, SMS is inescapable. It would be better if two-factor authentication used a less hackable alternative, but at the moment SMS is the widespread vector of corporate choice. We consumers don't actually get to choose to dump it until they do. A switch is apparently happening very slowly behind the scenes in the form of RCS, which I don't even know if my aged phone supports. In the meantime, Signal becomes the "another messaging app" we began with - and historically, diminished convenience has been one of the biggest blocks to widespread adoption of privacy-enhancing technologies.

Signal's decision raises the possibility that we are heading into a time where texting people becomes far more difficult. It may become like the early days, when you could only text people using the same phone company as you - for example, Apple has yet to adopt RCS. Every new contact will have to start with a negotiation by email or phone: how do I text you? In *addition* to everything else.

The Internet isn't splintering (yet); email may be despised, but every service remains interoperable. But the mobile world looks like breaking into silos. I have family members who don't understand why they can't send me iMessages or FaceTime me (no iPhone?), and friends I can't message unless I want to adopt WhatsApp or Telegram (groan - another messaging app?).

Signal may well be right that this move is a win for security, privacy, and user clarity. But for communication? In *this* house, it's a frustrating regression.

Illustrations: Midjourney's rendering of "railway signal tracks crossing",

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Stories about the border wars between cyberspace and real life are posted occasionally during the week at the net.wars Pinboard - or follow on Twitter.

Posted by Wendy M. Grossman at 10:59 AM | Permalink | Comments (0) | TrackBacks (0)

Bad ideas never die.

In particular, bad ideas in Internet policy never die. Partly, it's a newcomer problem. In the 1990s, one manifestation of this was that every newly-connected media outlet would soon run the story warning readers not to open an email with a particular subject line - for example, Join the Crew - because it would instantly infect your computer. These were virus hoaxes. At the time, emails were all plain text, and infection on opening an email was a technical impossibility. (Would that it still were.) This did end because the technology changed.

Still with us, though, are repeated calls to end online anonymity. It doesn't matter who it was this week, but there was a professorial tweet: social media should require proof of identity. This despite decades of experience and research that show that often the worst online behavior comes from people operating under their own well-known, real-world identity, and that many people who use anonymity really need it. And I do mean decades: it's 30 years since Lee Sproull and Sara Kiesler published their study of human behavior on corporate mailing lists.

This week, Konstantinos Komaitis, a senior director at the Internet Society, and 28 other Internet experts and academics sent a letter to the European Commission urging it to abandon possibly imminent proposals to require content providers such as Google and Facebook to pay "infrastructure fees" to telecommunications companies. The letter warns, as you'd expect, that bringing in such feeds upends the network neutrality rules in place in many parts of the world, including the EU, where they became law in the 2015 Open Internet Regulation.

Among prior attempts, Komaitis highlights similar proposals from 2012, but he could have as easily pointed to 2005, when the then CEO of AT&T, Ed Whitacre, said he was tired of big Internet sites using "my pipes" "for free". At the time, network neutrality was being hotly disputed.

The Internet community has long distrusted telcos. First, because the pioneers still remember their hostility to the nascent Internet and, as they will remind you at any mention of the International Telecommunications Union, because during the telcos' decades of monopoly were also decades of stagnation. A small sample of the workarounds and rule-breaking Internet founders had to adopt in Britain alone was presented at an event in 2013 that featured notable contributors Peter Kirstein, Roger Scantlebury, and Vint Cerf.

Of course, we all know what's happened since then: scrappy little Internet startups became Big Tech, and now everyone wants a piece of their wealth - governments, through taxation and telcos through changing the entire business model.

Until the EU's proposals surfaced last year, it was possible to think that this particular bad idea had finally died of old age. AT&T has changed CEOs a couple of times, and for a while in there it was owner of Time-Warner, which has its own streaming products. The fundamental issue is that the Internet infrastructure has grown up as a sort-of cooperative, in which everyone pays for their own connections and freely exchanges data with peers. In the world the telcos - and the postal services - live in, senders pay for carriage and intermediate carriers get a slice ("settlement"). Small wonder the telcos want to see that world return. (They shouldn't have been so dismissive at the beginning.)

EU telcos have been tilting at this particular wind turbine for a long time; in 2012, the European Telecommunications Network Operators Association (ETNO) called for settlement as part of a larger proposal to turn Internet governance over to the International Telecommunications Union. A contemporaneous 2012 presentation by analyst Falk von Bornstaedt argued that "sending party network pays" is the necessary future in order to provide quality-of-service guarantees.

The current EU call for this change is backed by Duetsche Telekom, Orange, Telefonica, and 13 other telcos. They have a new excuse: the energy crisis and plans for combating climate change mean they need Big Tech to share the costs of rolling out 5G and fiber optic cabling. More than half of global network traffic, they argue, is attributable to just six companies: Google, Facebook/Meta, Netflix, Apple, Amazon, and Microsoft.

It is certainly true that the all-you-can-eat model of Internet connection encourages some wastefulness such as ubiquitous Facebook trackers or constantly-connected subscription office software. Moving to "the metaverse", as Meta has $70 billion worth of hope that you will, will make this exponentially worse.

On the other hand, consider the truly undesirable consequences of changing the business model. The companies paying the telcos extra for carriage will expect in return to have their traffic prioritized. That in turn will disadvantage their competitors who don't have either that financial burden or that privileged access. Soon, what's left of the open Internet would be even more of an oligopoly, particularly with respect to high-bandwidth applications like video or virtual worlds, where network lag is the enemy of tolerable quality.

In a column (PDF), lays out the issues quite clearly and warns: 1) we may not have the tools to understand the consequences of such a change; and 2) we might not be able to unwind it if we regret it later, particularly if these companies continue to merge into even bigger and more predatory giants.

Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Stories about the border wars between cyberspace and real life are posted occasionally during the week at the net.wars Pinboard - or follow on Twitter.

Posted by Wendy M. Grossman at 12:47 PM | Permalink | Comments (0) | TrackBacks (0)