Carney can’t save us from American technology

The government shouldn’t bully the media. The government shouldn’t punish them for political speech. And the government definitely shouldn’t veil threats against political speech by offering to do this “the easy way or the hard way.” No country that values free speech would ever do that… right?

Hold that thought.

American far-right activist Charlie Kirk was assassinated last September. Late-night talk show host Jimmy Kimmel said in his opening monologue after the assassination: “We hit some new lows over the weekend with the MAGA gang desperately trying to characterize [the] kid who murdered Charlie Kirk as anything other than one of them and doing everything they can to score political points from it.”

Brendan Carr, chair of the U.S. Federal Communications Commission and the United States' number one censor, couldn’t help but do some censoring. Shortly after Kimmel’s monologue, Carr said that ABC, Kimmel’s broadcaster, could “find ways to change conduct and take action” on him, or there would “be additional work for the FCC ahead.” He added that the FCC could do this "the easy way or the hard way.” The threat followed conservatives’ calls for retribution against those critical of Kirk (including Kimmel).

This is naked intimidation from the U.S. telecommunications regulator against ABC, its owner Disney and ABC-affiliated local broadcasters. 

Carr’s argument was that Kimmel intentionally misled the public into believing that Tyler Robinson, the accused assassin, was a right-wing Trump supporter. Robinson’s politics were the subject of intense partisan debate.

Did Kimmel violate any rules? We’ll never know, because the FCC never had to decide. Shortly after Carr’s statements, two of the largest operators of ABC-affiliated stations announced that they would be pulling Kimmel from all local stations; ABC later indefinitely suspended the show.

Even without official action, the threat was enough.

The FCC’s influence might not seem relevant here in Canada, but many of the services we use at UBC are American. Using Canvas, Workday and Microsoft 365 is required to study at UBC. Students use AI tools created by OpenAI, Google and Anthropic on computers running Windows and macOS by accessing the internet through Google Chrome. Outside of class, we watch TV made by American media companies, scroll through Instagram and spend hours on YouTube. Every single one of these services is American.

These companies benefit from remaining reasonably politically neutral. A MAGA graphic on the Google homepage would alienate a massive chunk of their userbase. Simultaneously, these companies also benefit from remaining in the good graces of the U.S. government — a government that currently isn’t afraid to leverage its power over a company to get what it wants.

Consider another situation involving Carr. Earlier this year, CNN published the details of a peace deal between the U.S. and Iran. Carr posted on Twitter (now X) that their reporting was a hoax, saying it was “time for change at CNN.” 

CNN is owned by Warner Bros. Discovery; when Carr made his post, it was all but confirmed that WBD would be bought by Paramount Skydance. The FCC has not yet approved the merger, but Carr said in March that he thought it was a “good deal” and that it would likely get approved quickly.

Paramount Skydance is owned by Larry and David Ellison. Larry Ellison is the eighth-richest person in the world; David Ellison is his son. Both are right-wing Trump supporters. Paramount Skydance also owns CBS. 

The FCC doesn’t have direct regulatory power over CNN — the network operates over cable while the FCC regulates broadcast TV — but it does have power over CBS, a broadcaster that has caved to Carr before and is led by Trump allies.

Carr calls for change at CNN while applying regulatory pressure to CBS; both are now part of a company owned by a billionaire Trump supporter. 

See how this works? There are many pathways for the government to pressure companies. Some are obvious, some less so. 

Defense deals are a prominent example. Google and Amazon both have deals with the U.S. Department of Defense, as do many other AI providers. A coordinated Trump administration could pressure these companies to change their consumer-focused software or risk losing lucrative defense deals.

This has somewhat happened with Anthropic. In May, the Pentagon announced deals with most major AI companies, including Google and OpenAI. It did not sign a deal with Anthropic, declaring it a supply chain risk.

This was due to Anthropic’s two red lines, which they refused to cross: lethal autonomous weaponry and mass domestic surveillance. This decision also comes in part from the administration’s tirade against “woke” values and DEI; the administration considers Anthropic and its products to be woke. (Notably, Anthropic’s red lines match existing DoD directives on AI use.)

In June, the American government issued an export control directive ordering Anthropic to suspend foreign nationals’ access to its powerful Mythos 5 and Fable 5 models.

Again, it’s not clear why exactly this happened, but Axios reported that “Anthropic has not done a great job at trying to speak to the administration and appreciate the ideological differences,” according to a source familiar with the government’s thinking.

In both instances, at best, the U.S. government lashed out against a major American company for dubious reasons; at worst, Anthropic’s exclusion was motivated by the government’s political disagreements with the company.

It’s impossible to speculate about all of the ways the U.S. administration could pressure the companies that create the software we use at UBC. The government has virtually endless levers to pull, and the bar for these companies to stand up to the Trump administration is very high. Remember, they care about money. Not you. 

If the Trump administration can take away more than a company would lose by alienating customers, caving to the administration is just business. Even if they do stand up to the Trump administration, they’ll face very costly legal battles, potentially causing students and universities to lose access to their tools and move to competitors.

Since the U.S. government is not afraid to use its power over these companies for political gain, how do we protect ourselves?

The answer, unsurprisingly, is that we need a non-American collection of software that we can use. In software development, this is referred to as a “tech stack.” Technologies are stacked to build a program, ranging from the server a program is hosted on to the operating system of the device you use to run it to the program itself. I’m using “tech stack” more broadly here to encompass everything that would be needed to run software as independently from the United States as possible.

The idea of a Canadian tech stack gets thrown around every so often and has been gaining traction since Trump’s election in 2024. A Canadian tech stack would encompass the software that we use every day and the infrastructure that supports it: social media, streaming services, educational software, AI tools, cloud storage, servers, data centres and more.

Prime Minister Mark Carney’s AI for All plan is part of a possible Canadian tech stack. The plan has three core parts: building trust, creating opportunities and reinforcing Canadian sovereignty. This would include building “compute, cloud, connectivity, data and talent” for sovereign Canadian AI “on Canadian terms.” 

Carney’s plan would definitely help to mitigate the influence of the U.S. government on Canadians and other non-Americans who use AI tools. Problem solved!

Except… What happens when the same forces that have taken hold of Washington get into Ottawa? One only needs to look at Pierre Poilievre, Danielle Smith and the newly-elected leader of the B.C. Conservatives, Kerry-Lynne Findlay, to find different flavours of politics adjacent to the MAGA movement. Or maybe Carney, whose government recently introduced a bill banning social media for minors (which, as I’ve discussed, is bad), would want to exert control over Canadian technology for political gain.

Maybe none of that is an issue. Maybe neither Carney nor the Conservatives will use their power for political gain. But maybe they will. 

So now what? The United States is off the table, and if we can’t trust any country to protect our online rights, what do we do?

If a Canadian tech stack (or one from any other country) is just as susceptible to the authoritarian political pressures seen in the U.S., we need a global tech stack. Canada should support digital sovereignty — so should the European Union, Mexico and every other country you can think of. You should be able to use software from the United States or Canada, but if those countries become political risks, you should be able to move to comparable alternatives based elsewhere. 

If you don’t want to use ChatGPT or Google Gemini, you should have a Canadian option; if you don’t want to use that, you should have a German option, a Japanese option and a Mexican option. 

This is, admittedly, a pipe dream. The world just isn’t set up for distributed technology development. AI talent, as with most software development, is concentrated in Silicon Valley and the larger Bay Area. But we’re getting closer. The Canadian government supports sovereign AI. The European Commission has joined W, a social network that is interoperable with Bluesky and the open AT Protocol. The world is recognizing that the technology we use shouldn’t be based in the United States. It shouldn’t be based in any single jurisdiction.

If Canada follows in America’s steps and starts chilling Canadians’ speech, you should still have options for your digital life. Right now, we’re stuck with a bunch of American companies cozying up to a fascist. Even if you never step foot across the border, you’re paying the price.