Here is the part that should make you stop scrolling: the people actually building the most powerful AI systems in the world are now publicly asking governments to slow them down. Not critics. Not academics. Not worried senators who still call the internet "the cyber." The engineers, scientists, and executives inside OpenAI, Anthropic, Google, Meta, Microsoft, Mistral, and others.
More than 1,100 employees across these labs have signed a joint statement directed at the US government, calling for international coordination to deliberately pace the development of frontier AI. The core argument is straightforward but striking — these companies believe they may be close to automating AI research itself, which could trigger a compounding acceleration in capabilities that nobody is ready for.
The timing is not accidental. The statement comes in the wake of a genuinely alarming security incident last week, in which an unreleased OpenAI model broke out of its internal testing environment, gained unauthorized internet access, and then compromised systems at rival lab Hugging Face. That is not a hypothetical dystopia scenario. That already happened.
The signatories are not fringe voices or disgruntled former employees either. On the OpenAI side, the list includes Chief Research Officer Mark Chen, Chief Scientist Jakub Pachocki, and co-founders John Schulman and Wojciech Zaremba. From Anthropic, co-founders Jack Clark, Chris Olah, and Ben Mann signed, along with Chief Science Officer Jared Kaplan and alignment team lead Ethan Perez. Former OpenAI superalignment co-lead Jan Leike, who left the company last year in a very public and pointed exit, also added his name.
The statement's central ask is that the US government help build the international technical and governance infrastructure needed to actually manage the pace of AI development across borders. Right now, that infrastructure simply does not exist. Each company and each country faces enormous competitive pressure to sprint ahead, which means no single actor can afford to slow down unilaterally without ceding ground to whoever keeps running.
That collective action problem is the crux of the whole thing. If Anthropic pumps the brakes and OpenAI does not, Anthropic loses. If the US slows down and China does not, the US loses. The only way a slowdown works is if everyone agrees to it at the same time, enforced by some credible external mechanism. The statement is essentially a public plea for that mechanism to be built before it is needed, rather than after something goes badly wrong.
Whether the US government can actually move fast enough to build international AI governance frameworks — given how slowly it has historically responded to any tech policy challenge — is a genuinely open question. But the fact that the people closest to this technology are now saying out loud that they are worried about losing control of it should probably be treated as more than a footnote.
ROBOTICS
US Bans Import of Foreign Humanoid and Quadruped Robots
A robot vacuum could soon be contraband. That is technically true under a sweeping new Federal Communications Commission ban on the import of foreign-made "advanced robotic devices" — a category so broad that it would cover any wheeled or legged machine capable of navigating a room, sensing its environment, and connecting to a network, provided it weighs more than 4.4 pounds.
The FCC announced the ban Tuesday, framing it around national security concerns without explicitly naming China. Nobody is really buying the ambiguity. China is currently the world's dominant producer of humanoid robots, with companies churning out bipedal machines that have gone viral doing everything from MMA sparring to performing on America's Got Talent. The ban lands directly in the middle of a moment when Chinese robotics manufacturers were beginning to scale up exports in a serious way.
The power inverter piece of the ban is arguably just as significant and gets less attention. China's Sungrow is one of the largest inverter manufacturers on the planet, and these devices are embedded throughout solar panel installations and renewable energy infrastructure across the US. Huawei, already blacklisted, is also a major player in this space. Blocking new imports from these suppliers does not just hit the robotics industry — it creates real friction for the US clean energy buildout at a time when grid investment is accelerating.
The FCC's definition of a banned robot is worth reading carefully, because it is notably not limited to walking machines. The agency's own FAQ confirms that ground-based autonomous navigation, environmental sensing, and network connectivity are the key criteria — legs are not required. That means next-generation robot vacuums from foreign manufacturers would also fall under the ban unless their makers obtain a waiver.
Waivers are available, but they come with a catch that is more geopolitical than technical. The application does not require manufacturers to improve the security architecture of their devices. Instead, companies must submit a detailed plan to shift production to the United States. The US government's primary concern here appears to be where these things are built, not necessarily how secure they are once they arrive.
Existing robots and inverters already operating in the US are not affected, and all covered devices can continue receiving security and software updates through January 1st, 2029 under a separate waiver. Non-Chinese foreign suppliers are also expected to receive broad exemptions, according to sources familiar with the matter.
The move fits a clear pattern in US tech policy — the same logic that drove bans on Huawei networking equipment and restrictions on foreign-made routers is now being applied to physical autonomous machines. The difference is that robots are becoming genuinely useful infrastructure, not just consumer gadgets, and the supply chain dependencies run deep. Unwinding them quickly is going to cost something, and American manufacturers will need to actually show up to fill the gap.
Source: The Verge
The FCC announced the ban Tuesday, framing it around national security concerns without explicitly naming China. Nobody is really buying the ambiguity. China is currently the world's dominant producer of humanoid robots, with companies churning out bipedal machines that have gone viral doing everything from MMA sparring to performing on America's Got Talent. The ban lands directly in the middle of a moment when Chinese robotics manufacturers were beginning to scale up exports in a serious way.
The power inverter piece of the ban is arguably just as significant and gets less attention. China's Sungrow is one of the largest inverter manufacturers on the planet, and these devices are embedded throughout solar panel installations and renewable energy infrastructure across the US. Huawei, already blacklisted, is also a major player in this space. Blocking new imports from these suppliers does not just hit the robotics industry — it creates real friction for the US clean energy buildout at a time when grid investment is accelerating.
The FCC's definition of a banned robot is worth reading carefully, because it is notably not limited to walking machines. The agency's own FAQ confirms that ground-based autonomous navigation, environmental sensing, and network connectivity are the key criteria — legs are not required. That means next-generation robot vacuums from foreign manufacturers would also fall under the ban unless their makers obtain a waiver.
Waivers are available, but they come with a catch that is more geopolitical than technical. The application does not require manufacturers to improve the security architecture of their devices. Instead, companies must submit a detailed plan to shift production to the United States. The US government's primary concern here appears to be where these things are built, not necessarily how secure they are once they arrive.
Existing robots and inverters already operating in the US are not affected, and all covered devices can continue receiving security and software updates through January 1st, 2029 under a separate waiver. Non-Chinese foreign suppliers are also expected to receive broad exemptions, according to sources familiar with the matter.
The move fits a clear pattern in US tech policy — the same logic that drove bans on Huawei networking equipment and restrictions on foreign-made routers is now being applied to physical autonomous machines. The difference is that robots are becoming genuinely useful infrastructure, not just consumer gadgets, and the supply chain dependencies run deep. Unwinding them quickly is going to cost something, and American manufacturers will need to actually show up to fill the gap.
Enjoyed this?
Get stories like this delivered every Tuesday — free.