Washington Just Declared AI Untouchable — And the Men Building It Are the Ones Sounding the Alarm
(SeaPRwire) - By: Julian Holbrooke There is a peculiar inversion unfolding in Washington. The people who build artificial intelligence are begging for restraint. The man who governs them is promising none. On Saturday, Donald Trump took to Truth Social to vow that his administration "will not in any way hinder or stifle the Growth of this incredible Industry." He went further, announcing an "AI Force" modeled on his first-term Space Force, and promising an AI czar for whom "only High I.Q. individuals need apply." This is not technology policy. It is campaign theater dressed in the language of industrial strategy, and it lands at precisely the moment when the industry's own leadership has lost its nerve. When Dario Amodei, Sam Altman, Elon Musk, and Demis Hassabis all signal that the systems they are shipping may be outpacing human control, a president dismissing those fears as a "hoax" is not making an argument. He is making a bet with other people's downside. Read the official text closely and the architecture of the message becomes clear. Trump frames AI as potentially more consequential than the Industrial Revolution and the internet, and floats a figure of 25% of US GDP. He casts political opponents as agents of "decimation, or destruction," blaming "Radical Left Dumocrats" for what he calls a series of hoaxes. Abuses of the technology, he argues, can simply be handled through the justice system after the fact. Note what is absent. There is no detail on the AI Force's composition, its powers, its budget, or where it would sit inside the federal government. The Space Force comparison is doing heavy lifting here, but that analogy collapses under scrutiny. Space Force was established by Congress in 2019 to protect defined physical assets, satellites among them. AI has no equivalent perimeter. You cannot garrison a gradient descent. Now set that against the second half of the record. Amodei warned last week that rapidly advancing systems could escape human control, be turned toward cyberattacks and bioterrorism, and cause serious economic disruption. Evan Hubinger, who leads alignment science at Anthropic, put a number on the table that should chill any serious policymaker: a greater than 10% chance that AI could "kill all humans" within the decade. One can dispute the probability estimate. Serious people do. But the geopolitical subtext of Trump's position is not really about safety philosophy at all. It is about China. "We are leading China, and the rest of the World, and I intend to keep it that way," he wrote. In that framing, every safety constraint becomes a unilateral handicap, and every warning from a lab CEO becomes indistinguishable from sabotage. That is a convenient logic for incumbents who want regulatory clearance, and a dangerous one for everyone else. The gravest detail in this story is also the quietest. Bloomberg reported earlier this week that flawed military intelligence, outdated imagery, and overreliance on AI contributed to a February strike on an Iranian school that killed more than 150 people, at least 123 of them children. That is not a thought experiment from an alignment paper. It is the technology's current failure mode, deployed, with a body count. The geopolitical pendulum is swinging toward acceleration at exactly the moment the evidence argues for brakes, and history suggests the bill for that miscalculation will arrive long before Washington admits it placed the wrong wager. Author bio: Julian Holbrooke, an overseas international relations analyst who frequently contributes to major European daily newspapers on technology statecraft and transatlantic security affairs.
More


