From Fortune · Darlene Superville · · 2m
Trump shrugs off AI slowdown calls as Obama tells Democrats to run on it
AI safety just became an election issue, and the two parties are already picking very different fights.
In brief
AI safety just became an election issue, and the two parties are already picking very different fights. Washington now agrees AI needs guardrails and still cannot say what they are, so the race with China keeps setting the pace. Originally reported by Fortune.
Whoever wins AI wins
Trump batted away the idea of slowing AI down, framing the whole question as a race the US is currently winning against China.
“But whoever wins with AI wins.”
Guardrails, but unnamed
He allowed that some regulation is needed, then offered no specifics and blamed the warnings on unidentified "negative forces" raising things that won't happen.
“you have a lot of negative forces that are bringing it up that … shouldn't be bringing it up”
The CEOs asked for brakes
Anthropic's Dario Amodei said the industry should slow development so safety work can catch up. Elon Musk and Sam Altman publicly agreed with him.
Xi is coming in September
The Chinese president visits the US later this month, and AI is expected to be on the table with Trump.
Data centers are the local fight
Big companies are seeding small towns with computing warehouses. Voters in both parties are angry about lost farmland, water use and higher power bills, and many towns are trying to block them.
Obama's pitch to Democrats
Run on AI. He wants a stated plan for safety, for protecting kids, and a concrete answer on which jobs get displaced and where.
“What does it mean if there's going to be job displacement. Where is that going to hit? How are we going to respond?”
Congress agrees on nothing specific
Speaker Mike Johnson wants one big meeting with industry leaders. Jeffries wants decisive action now. Neither named a single concrete step.
“But we don't need everybody to panic right now.”
An insider wants a truce
Jacob Coxon, who quit Anthropic over safety fears, argues the US and China should agree to slow down together rather than sprint.
“You're building a human-level or superhuman-level mind, without understanding what it wants, or thinks”
Sacks throws it back
White House tech adviser David Sacks told the AI bosses asking for rules that they set the pace themselves and can stop whenever they like.
“The easiest way not to build superintelligence is for you to agree not to build it.”
Altman's own warning
In a Fortune interview, the OpenAI chief said safety standards are not ready for a big capability push and called AI beyond human control absolutely possible.
“not at a place”
What to remember
Washington now agrees AI needs guardrails and still cannot say what they are, so the race with China keeps setting the pace.
