US Blocks G20 AI Regulation: What This Means for Global AI Safety
The US Just Told the World: Don't You Dare Try to Control AI
I need you to read this next sentence slowly, because I don't think it's gotten the attention it deserves. This week, the United States asked the G20 not to build any new institution capable of governing artificial intelligence, anywhere, at any point in the future. Not "we want lighter rules". Not "let's regulate this differently". The actual position on the table is that no body with the authority to write global AI rules should exist at all.
I've spent a lot of time talking about the Great Filter, the Fermi Paradox, and whether we're racing toward building something we can't put back in the box. This story isn't a thought experiment anymore. This is the exact dynamic I've been warning you about, playing out in real time, in an actual policy meeting, with real names attached to it.
What Actually Happened This Week
A two-day G20 meeting opened in North Carolina, and a White House official laid out the American position before it even started: no new international institutions to govern AI. Commerce Secretary Howard Lutnick and White House technology adviser Michael Kratsios are hosting the session. This is the opening move in a series of negotiations that runs all the way to the G20 leaders' summit in Miami this December, which means the position set this week is exactly what the rest of the world will spend three months arguing against.
Now here's the part I want you to sit with. The argument being made isn't about specific rules Washington wants. It's procedural, and it's procedural on purpose, because that's what makes it effective. You can't negotiate the details of a rule that's never allowed to exist in the first place. That's not a policy position. That's a strategy to make sure the conversation never gets far enough to produce one.
Look at Who's in the Room
I want you to pay close attention to the guest list, because it tells you everything about how this administration actually sees this issue. Sam Altman and Jensen Huang are attending in person. Elon Musk is joining by video. Let that sink in. The people whose companies stand to profit most from AI development moving forward with zero binding international oversight are sitting in the room while their government argues that no oversight body should ever exist. I don't need to spell out the conflict of interest here. It's not subtle, and it's not hidden. It's happening in plain sight, at an actual intergovernmental meeting.
This is precisely what I meant when I talked about the warning that the real danger of AI isn't some evil, scheming machine. It's a system becoming powerful enough that even small misalignments between what we want and what gets built become impossible to correct in time. You cannot correct what you've made structurally impossible to govern. That's not a technical risk anymore. That's a choice being actively defended by the people with the most to gain from making it.
Why "No New Rules" Is More Dangerous Than "Wrong Rules"
I think a lot of people will read this story and shrug, because "government blocks new regulation", sounds like the kind of headline we've all grown numb to. I need you to understand why this specific instance is different, and worse.
Europe already built what the US is arguing against. The AI Act is in force. The Digital Security Act is actively being enforced against ChatGPT and other major platforms. The Cyber Resilience Act took effect this very month. Europe isn't hypothesizing about oversight, it's running it. And a G20 consensus against international AI governance wouldn't undo any of that directly, but it would do something almost as damaging; it would make Europe's actual, working regulatory framework look like an isolated outlier instead of a model other nations might reasonably adopt. That's the quiet, structural goal here, not destroying existing regulation, but making sure it never spreads.
Here's the detail that should genuinely disturb you. The chair of the Financial Stability Board told G20 finance ministers this same week that AI-driven cyber risk is now the single most immediate threat to global financial stability. That warning came through exactly the kind of multilateral institution the US is arguing should never be allowed to exist for AI specifically. And separately, a United Nations panel has warned that AI development is now outpacing both our scientific understanding of it and every government's ability to write policy fast enough to keep up. That's the case for new oversight institutions, made by an existing international body, in the same week the US is arguing no such institution should be built. The warning and the refusal to act on it are happening simultaneously, in public, and almost nobody is connecting the two.
This Is Not a Neutral Policy Debate
I want to be fair here, because there is a real, honestly argued counterpoint. Institutions built right now would be designed around a technology nobody fully understands yet, staffed by people appointed for a landscape that will look completely different in five years, and historically slow to adapt once they exist. That's a legitimate concern, and I'm not going to pretend it isn't.
But there's also a much less flattering truth sitting underneath this argument, one that almost nobody at that meeting will say out loud. A country that hosts most of the world's leading AI model developers has objectively less to gain from international oversight than countries that host none. Developing economies, the nations that neither build these models nor have the domestic capacity to regulate them effectively, are the ones with the most to gain from international standards, and the least power to demand them. This isn't a debate between equally matched interests. It's a negotiation where the country writing the rulebook happens to be the same country whose companies benefit most from there being no rulebook at all.
Why This Connects Directly to Everything I've Warned You About
I've written about the Great Filter hypothesis, the idea that advanced civilizations might not survive the technology they create to help them. I've written about tech billionaires calling AI something close to a god behind closed doors. This story is the missing link between those ideas and the actual, present-day world we're living in. It's not a hypothetical anymore. It's a real vote happening at a real summit table, and the argument being pressed by the most powerful nation on earth is that no global check should ever be allowed to exist on the exact technology that serious physicists and philosophers think might be the thing that ends us.
The G20 works by consensus. That means a communiqué that says nothing about AI governance counts as a win for the argument that nothing needs to be governed. Between now and December, the real question is whether any other nation is willing to spend the diplomatic capital it would take to write a single sentence the United States has already said it won't accept.
Where I Land
I don't think this is a partisan issue, and I'm not writing this to tell you who to vote for or which country's approach is perfect. I'm telling you that the single most consequential technology humanity has ever built is currently being kept deliberately outside the reach of any global governing body, by the same nation whose companies profit the most from that exact outcome, with those same companies' leaders sitting in the room while it happens. That is not a technical detail. That is not bureaucratic noise. That is a decision, made in public, this week, about whether anyone on Earth gets a real say in how this technology develops before it's too powerful to have that conversation at all.
We are, as I've said before, the first civilization actually having this conversation out loud, in public, while the technology is still being built. This week is a test of whether we actually mean that, or whether we let the people with the most to gain talk everyone else out of even trying.
Michael Routhier is the founder of Tech 4 Grown-Ups, providing honest, unfiltered digital literacy for adults 55+, and host of The Virtuous Machine, exploring the ethics and human cost of AI. Read by tech-curious readers in 50+ countries. Explore more at tech4grownups.com.
Comments