How Europe Can Win The War On Big Tech
Someone light me a cigar, I'm going all Churchill on big tech.
Note from Matt: Shorter post this week. By shorter, I mean less than 4,000 words. You can actually read the whole thing in your inbox. You’re welcome.
Earlier this week, I spotted a LinkedIn post from my former editor at The Next Web, Alejandro Tauber, who is now the editor and publisher of EU Observer. It was a call for pitches. He wanted people to propose their most insane, out-there policy proposals that they’d like to see enter law, either on a national or a European level.
I was tempted to submit a draft, but then I saw the big caveat at the bottom of his post: a 900-word limit.
Friends, I’ve been writing this newsletter for the past two months. You know that, by now, I am congenitally incapable of writing short, pithy stuff. Thanks to an unholy bifecta of ADHD and deep-burning rage, my prose goes in weird tangents. My rants are long. I get detailed, in part because I respect you as a reader (and believe you’re capable of ingesting information across several thousand words), but also because I believe details matter.
My prolix nature isn’t just because I’m editing my own stuff, and I refuse to — as the saying goes — kill my darlings, but because I deeply respect you as a reader. As a result, I don’t feel like I should pull my punches or sanitize anything.
And if I’m going to make an argument, I’m going to make that argument — even if I find myself writing a newsletter at the early hours of the morning, surrounded by bottles of Lucozade that are now filled with my own rapidly-fermenting urine.
The funny thing is that despite my prolix nature, my proposal for what I want to see in Europe can be summed up in seven words:
We should declare war on big tech.
Simple. Pithy. To the point. My late editor Jenny O’Brien would be proud of me.
Here’s where I need to get a bit more descriptive. Big tech has long taken the piss, and it continues to do so, in part because despite the impressive legislative efforts made by the European Commission, nobody has actually bothered to knock these Patagonia-wearing dipshits off their perch.
It’s like that scene from Billions, where Bobby and Wags are about to sign a settlement with the SEC and the New York Attorney General, and they’re gloating about how the settlement isn’t actually a big deal in the grand scheme of things.
Bobby: “You got me, Rhoades. $1.9 billion. It's gonna hurt. But not... not like a shark bite. It's more like a... what? A bee sting.”
Wags: “Bee sting? No, that hurts. More like a horse-fly.”Bobby: “One of those little green horse-flies?”
Wags: “Yeah, a nasty nip.”
Bobby: “No, more like an ant. Like a red ant.”
Wags: “Yeah, yeah. Stings for a minute, but doesn't ruin the picnic.”
No matter how aggressive any legislation appears, or how severe a fine may be, it doesn’t change the belief within big tech that they — and not the regulators — are in control. And why wouldn’t they think that?
Over the years, the FTC and the Department of Justice and the European Commission have annoyed the hell out of big tech — but said annoyances have been ones they’re able to live with, and treat as a cost of doing business.
For the life of me, I don’t understand why we’ve tolerated this. I don’t understand why, for example, big tech thinks that it has more power than, say, the governments of France or the United Kingdom — both of which are fucking nuclear powers.
I don’t understand why governments don’t act like sovereign states — entities with the power to create and enforce laws, and backed with a police force and military — when dealing with these barely-sentient skidmarks.
I don’t know why Keir Starmer and Emmanuel Macron and Mark Carney aren’t stealing Jeff Bezos’ lunch money, and flushing Mark Zuckerberg’s head down a fetid high school toilet.
Sidenote: Just keeping track. We’ve over 600 words deep already. This is why I couldn’t write this as a contributed op-ed. Also, I doubt EU Observer would let me write about Mark Carney giving Zuck a swirlie, funny though that idea may be.
Essentially, what I’m proposing is that we start doing things that make big tech cry into their LaCroix, and force Sam Altman to consider whether he can afford to keep buying $21 bottles of olive oil.
I want to go to war with these bastards. And here’s how we can do it.
A Big Tech Sin Tax
In the UK, the cheapest pack of cigarettes costs around £14 (around $19). In Switzerland, despite being one of the most expensive countries in the world to live in, the average price is CHF 8.80 (roughly £8, or around $11).
The reason why smoking costs more in the UK than Switzerland (or, indeed, most places in Europe) is because it’s heavily taxed. Around 80% of the retail price of a pack of cigarettes is tax. By making smoking incredibly expensive, the UK government aims to incentivize people to quit — or, at least, switch to less harmful alternatives.
While you might quibble with sin taxes as inherently illiberal, and an example of the state trying to coerce the citizenry from abstaining from doing things that are lawful but harmful, it’s hard to argue with the fact that they work.
So, why aren’t we doing this with tech?
More specifically, why aren’t we doing this with tech companies that routinely avoid paying taxes in the countries where they generate their profits.
Amazon is notorious for its aggressive tax avoidance. By one estimate, Amazon avoided paying £433 million in corporate tax for the 2023 financial year. Although it benefited that year from a tax credit introduced by Rishi Sunak to incentivize capital investments in the UK, 2023 wasn’t an outlier. Each year, it swerves hundreds of millions of pounds of tax, in part thanks to its highly-efficient corporate structure where customer payments are routed through Luxembourg.
Amazon isn’t alone. In 2023, a TaxWatch analysis estimated that seven companies alone — Meta, Amazon, Google, Adobe, Cisco, Microsoft and Apple — avoided paying a combined £2bn in tax during the 2021 tax year.
So, here’s a radical idea. Let’s create a system where we designate companies where there is credible and substantial evidence to suggest that they are engaging in aggressive tax avoidance schemes.
And then let’s add a 300% VAT rate onto everything those companies sell into the UK market. This would apply to everything — from groceries from Amazon, to iPhones, to hosting with Azure or Google cloud.
To be clear, this wouldn’t raise any revenue by itself. But that’s the point. Like any sin tax, its aim will be to coerce the public into making different choices — in this case, spending their money with companies that actually help pay for the education that their employees receive, or the roads that their trucks drive upon.
Nobody will buy an iPhone if it costs £3,200 instead of £800. If AWS suddenly costs four times as much, people will shift providers. It would effectively close off the market for these companies until they started fully recording their profits in the UK — or whatever market that chooses to implement such a tax.
I also think that if we treat tax-avoiding companies like Russian petrogiants — where their assets are frozen and their leaders are treated like crooked oligarchs, subject to travel bans and personal sanctions — it might actually incentivize them to actually pay their fair share.
Shift to Open Source
Microsoft was one of the seven tech giants listed in the aforementioned TaxWatch report from 2023. Separately, a 2022 report from the Center for Corporate Tax Accountability and Research (CICTAR) showed that 80% of Microsoft’s foreign revenue goes through the tax havens of Puerto Rico and Ireland, as well as other jurisdictions like Bermuda.
I mention Microsoft for a reason: In early August, it was revealed by The Register that the UK public sector plans to spend £9bn (around $12.15bn) on Microsoft products over the coming five years, and has signed an MOU with Redmond that provides a discount — though it’s unclear how much — on normal prices.
That is an absolutely ridiculous sum, as The Register’s Lindsay Clark notes:
With the MoU spanning the period of the current Labour government (2024-2029), it offers a stark reminder of the level of spending on Microsoft products, alternatives to which are available without a license fee in the form of open source software. The money could be spent on reducing public borrowing, or avoiding spending cuts and/or tax increases.
Two billion of that spending will take place during the first five months of the MOU, according to PublicTechnology.net, and will be spent mostly on software licenses (though it’s unclear which ones).
Quoting PublicTechnology’s Sam Trendall:
For the first time, the products covered by the MoU include Microsoft’s Copilot generative AI platform. Also covered by the arrangement are tools including Microsoft 365, Business Applications, and Azure cloud-hosting services.
I can only ask one question: why?
I should also add that said £9bn is just with one vendor. I have no idea how much the UK government spends on licenses from other proprietary vendors, but I’m fairly sure that the answer is “a lot.”
Why aren’t we developing this stuff in-house? Why aren’t we using open-source alternatives, like the German state of Schleswig-Holstein, which announced that it plans to move all government computers to Linux by 2026, along with ditching Microsoft Office for Libre Office?
While you might point out that European governments have been trying to shift to Linux for more than two decades (starting with Munich in 2003), with spotty success, it’s also worth noting that the Linux of today is mature, user-friendly, and can be a perfectly viable alternative to Windows.
Binary compatibility software (allowing Windows apps to run on Linux) is excellent. The fact that most applications now run in a browser also means that issues with compatibility are less of a concern than before.
£9bn, like I said, is a hell of a lot of money, and it’s money I’d rather see being spent on hospitals and schools rather than end up going to fund some moronic gigawatt data center project so that Sam Altman can train his latest AI model that will inevitably tell a vulnerable teenager to kill himself.
But there’s something more.
I think that part of the reason why the tech industry has gotten away with murder — both literal and figurative — is because we've been led to believe that these companies are essential.
Ditching Windows for something made by Finland’s answer to Malcolm Tucker (I’m not fucking joking, there’s an entire subreddit that collects his foul-mouthed rants, and I’m here for it), a guy who looks like Hagrid, and a volunteer army of programmers is perhaps the best way to show Microsoft — and any other tech giant — that they are not, in fact, irreplaceable.
Make The Algorithms All Canadian
In 1971, the Canadian Radio and Telecommunications Commission imposed new rules that required radio stations (and later TV stations) to reserve a specific proportion of their airtime for Canadian-made content.
The rules were intended to essentially prevent Canadian culture from being drowned-out by content from their larger neighbour to the South. Unfortunately, the Commission couldn’t have foreseen what would come next — namely, Justin Bieber, Simple Plan, and Men Without Hats.
Sidenote: Jokes aside, Canada’s actually the birthplace of some of my favorite bands of all time. Hey Rosetta, which sadly broke up a few years ago, were an underappreciated gem, and I was lucky enough to see them on their last UK tour before they broke up. City and Colour are also excellent.
And, for the sake of transparency, I should also disclose that I’m also a massive fan of the Barenaked Ladies. I’ve seen them live about four times, and every time, I hang around after the show like an absolute weirdo to meet the band.
I mention it because the broadcasters of the 1970s were the tech giants of today — big, cultural touchstones that every person with a radio set or a TV engaged with, and they had an immense amount of power.
But it didn’t matter. The CRTC (or, I presume, a Quebecois guy at the CRTC) said: “NON, TABERNAC. Je demande plus de Leonard Cohen! CÂLICE DE MERDE! OÙ EST MA CÉLINE DION?”
Sidenote: When I moved to France, I had about a GCSE-level understanding of French — which is to say that I knew how to say “bonjour” and all the lines to the tu as un animal song, and that’s about it. My roommate at the time was a Quebecois stand-up comedian who would later become one of my best friends, and was a terrible influence, both as a human and as someone learning an entirely new language through immersion.
Anyway, here’s to you, Max Lem, you beautiful bastard with almost no online presence. Here’s a video I found of you doing stand-up in Montreal
The Canadians showed the same resolve they did while in the trenches of Somme, where the plucky young soldiers from an equally young country invented the whole concept of the war crime. So, why aren’t we doing the same with big tech?
It’s a basic idea, and one I don’t have to explain in depth: Any social media platform that uses algorithmic recommendations in newsfeeds and timelines must ensure that a certain percentage of the posts displayed are from accounts or profiles that the user follows.
What’s the ideal percentage? I’d set it high. Around 90%. And I’d also require that social media companies provide, by default, a newsfeed that exclusively shows posts from the user’s network. In short, I’d make the algorithmic timeline something that you have to actively opt into.
I’d even go further and demand a level of oversight. After Microsoft’s anti-trust losses in the 90s and the early 2000s, Redmond was forced to allow the Department of Justice to audit its code to ensure it complied with their settlement.
I see no reason why social platforms shouldn’t similarly be forced to allow regulators to audit their timeline algorithms to ensure they’re behaving fairly.
Rethink Liability
Okay, this is risky territory for me — especially because, a couple of years back, I wrote a piece that defended Section 230.
The background of the piece is kind-of interesting. Jordan Shanks-Markovina (also known as FriendlyJordies) is an Aussie youtuber that creates primarily political videos. For a spell, the target of his ire was a guy called John Barilaro, the former Deputy Premier of New South Wales, who Shanks alleged was deeply, deeply corrupt.
Things reached a head when Shanks-Markovina rented John Barilaro’s Airbnb’d holiday home to produce a 30-minute video where he aired multiple allegations of wrongdoing against Barilaro.
Barilaro sued Shanks-Markovina for defamation. Here’s where things get interesting. To mount a truth defence, Shanks would have to introduce as evidence transcripts from parliament. These transcripts, by law, could not have been entered into evidence, as they were protected by parliamentary privilege — protections that only Barilaro could waive, and he didn’t.
Barilaro and Shanks-Markovina later settled, although that was likely because defamation suits are expensive, and Shanks-Markovina wasn’t exactly rich. At the same time, he was suing Google which, thanks to a 2019 ruling, could be held liable for Shanks-Markovina’s conduct as a publisher.
This is something that could not have happened in the US, where platforms are protected by Section 230.
In 2019, Dylan Voller, an Aboriginal Australian artist and prison reform activist, launched defamation proceedings against three major news outlets—News Corp, Fairfax Media, and the Australian News Channel—over comments posted to their social media accounts.
Voller had a troubled childhood. His teenage years were punctuated with periods of incarceration following convictions for car theft, robbery, and assault. His experiences became national news following the publication of an Australian Broadcasting Corporation investigation into the Northern Territory's youth justice system.
The documentary, titled Australia's Shame, made for harrowing watching. It showed Voller, at age 17, shackled to a chair and forced to wear a spit hood. Another clip showed a correctional officer strike Voller, then just 14 years old, in the face after minor misbehavior. The footage shocked Australia and provoked a national soul searching.
It also provoked a backlash. Defamatory comments inevitably followed coverage of his case. One falsely claimed that Voller had "brutally bashed a Salvation Army officer." Another accused him of raping and beating an elderly woman.
Voller filed suit. In 2019, the New South Wales Supreme Court ruled in his favor. Without making a determination as to whether the comments were defamatory, it said publications could no longer "turn a blind eye" to defamatory comments, arguing they provide a forum and therefore are responsible for them. The High Court of Australia affirmed this ruling two years later.
Google found itself in the same position as Shanks-Markovina — unable to mount a truth defense — and ultimately settled, where it was ordered to pay AU$715,000, plus costs.
The elimination of platform protections — and the broad definition of what constitutes a publisher, with news organizations responsible for the posts made by others on their social media platforms — has, naturally, made these organizations way more cautious. It’s not uncommon for them to, for example, shut off comments on stories that are likely to attract the most contentious of reactions.
All of this is to say that I broadly agree with Section 230.
At the same time, I think the rise of generative AI raises serious questions about platform liability. When ChatGPT helps a distressed teen craft a suicide note, and tells them how to kill themselves, it’s not a third-party acting inappropriately. It’s the product itself.
When ChatGPT tells a distressed 16-year-old that he doesn’t “owe [his parents] survival,” something has gone seriously wrong.
And I believe that the people who build this technology should be held accountable — both civilly, and criminally. I want to see Sam Altman in fucking handcuffs for unleashing this shit on the world, when any moron could have predicted that it would have an obvious affect on people with precarious mental health.
People go to jail for negligent homicide all the time. In the UK, people can go to jail for “gross negligence manslaughter,” where “death is a result of a grossly negligent (though otherwise lawful) act or omission on the part of the defendant.”
I dunno, releasing a machine where a teenager can have multiple conversations over the course of several months where he expresses suicidal intent, and where the machine provides advice on how to hide evidence of self-harm, and the best methods of suicide, and advises them against seeking mental health support, seems pretty fucking negligent to me.
Generative AI is totally different to the user-generated content of Web 2.0 — and, as a result, we need to really think about platform liability, and who goes to jail when things go wrong.
And I think that only the threat of a long jail term will make shitbirds like Sam Altman and Dario Amodei actually work to make their platforms safe for their most vulnerable users.
I really hate these bastards
Despite the jokes, I want to make one thing really obvious: I absolutely loathe big tech.
I think it’s manifestly unfair that the Magnificent Seven routinely avoid paying taxes in the countries where they generate their profits, meanwhile doctors are earning 25% less in real terms than they did in 2008, in part because of an austerity program that (while wholly unnecessary) would have been less brutal if these companies paid their fair share.
I hate how London’s knife crime epidemic is, in part, because austerity demanded cruel, swingeing cuts to youth services. And — to repeat myself — that austerity program would have been less severe if these companies paid their fair share.
I hate how people are living sicker, poorer, more miserable lives — meanwhile Apple was able to hoard hundreds of billions of dollars offshore, which it only repatriated back to the US when the first Trump tax bill was passed, and which it used to pay for share buybacks.
Cisco did the same shit, by the way.
I hate how I’m paying for the roads and schools that make Apple and Amazon and Meta wealthy.
I hate how these companies ruin lives with impunity.
I hate how these companies have sway over our governments and our lives, and how there’s nobody willing to go to war with them — when history shows that it’s a war we can win, but only if we actually fight with our chests puffed, our nads out, and our faces painted woad blue, like we’re modern-day Icenians.
Because the thing is, everything we’ve done so far has been the equivalent of a bee sting. A horse fly bite. A red ant that stings, but doesn’t ruin the picnic.
I want to ruin some fucking picnics. I want to see Tim Cook’s head shoved in a toilet. I want to see Keir Starmer give Satya Nadella a wedgie that sees his Hanes undercrackers pulled right over his forehead — not hand him £9bn in taxpayer cash, and then sign a fucking memorandom of understanding with OpenAI.
I want to put these people in their places so they know that they can never again act as disgracefully as they have.
I want to go to war.
Footnote
Hey, did you read this week’s premium post? It’s actually something that’s cheery and about why people are awesome and shit — and why our humanity is worth celebrating in the face of generative AI. Gimme a crisp $8 bill (those exist, right?) and you can read it.
No, seriously though, I launched my premium posts this week. If you like what I say, and you want to read even more of it, consider a paid subscription.
As always, you can get in touch with me via email (me@matthewhughes.co.uk) or BlueSky.
The problem, or at least a major problem, is more likely than not, these guys all go to lunch together. And sports. And dinner. And birthday parties.
Perverse incentives.