The Four Fights
What We Demand
Our AI Ground Rules are what we believe. But this is what we're fighting for specifically.
Don't take AI lobbying money
Politicians should answer to voters – not to Leading the Future and its network of AI super PACs. No contributions, no endorsements, no outside spending on their behalf.
The AI companies are pouring money into the 2026 elections. Their super PAC network, Leading the Future, has raised $140 million – including $50 million from venture capitalists Marc Andreessen and Ben Horowitz.
That kind of political investment expects a return.
Defend state AI protections
When your state passes AI safeguards, Washington shouldn't be allowed to wipe them out. Washington's rules should be a floor, not a ceiling.
The federal government keeps trying to rule that states cannot set their own AI laws – preempting the states (aka 'preemption').
In 2025, the Senate voted 99–1 to protect state AI laws. But the preemption push keeps coming back anyway – and until Washington passes real safeguards of its own, it has no business erasing the states'.
The AI companies aren't only building a better chatbot – that's not what they're spending hundreds of billions a year on data centers for.
Their stated goal is AI that matches or beats humans at every job. The AI CEOs admit they can't fully control their technology. So we may be headed for a world with machines smarter than humans that we can't control. Even if the machines do stay under human control, the companies are building for an economy that no longer needs human workers.
No one in power is planning for this. We need to make them.
AI companies should pay for mass automation: we need a real plan for jobs
Make the biggest AI companies and automation windfalls pay into an American wealth fund that every citizen owns a share of – paying a dividend like Alaska's, shoring up Social Security and Medicare, and putting a floor under every family if jobs disappear at scale.
That's one plan. If economists or the government have a stronger one, bring it. What we won't accept: retraining brochures, or no plan at all.
The people building AI predict mass job losses themselves – Anthropic CEO Dario Amodei warns unemployment could hit 10–20% within five years, with half of entry-level office jobs gone. Leading economists across the spectrum – more than a dozen Nobel laureates among them – warn the disruption could be bigger than the Industrial Revolution, on a fraction of the timeline. Washington's first move is a bipartisan push to count the job losses. A count is not a plan.
Don't build AI we can't control
No one should develop AI systems that could escape human control – and the world should be able to verify that no one is.
The companies aren't hiding the goal: smarter-than-human AI (aka “superintelligence”). Hundreds of the field's top scientists – including the CEOs building it – signed a statement calling it a risk on the scale of pandemics and nuclear war.
They know it's dangerous. They're building it anyway.
An international agreement
An international agreement, backed by a real verification regime, so that no one – no company, no country – builds AI that humanity can't control.
Want to go deeper?
We're not doing this alone
Other policy efforts we support:
Your Move
