Illinois Gov. JB Pritzker says very powerful AI needs strong safety rules—like how cars, airplanes, medicine, and other things that can hurt people have safety checks.

In an interview with ABC News, Pritzker said AI can be “dangerous on the level of nuclear weapons.” He was not saying every AI tool is a weapon. People use AI for helpful tasks, such as writing, translating languages, finding patterns in medical information, and helping students learn.

But Pritzker is worried about the most advanced AI systems, sometimes called frontier AI. These are the newest and most powerful systems. In the wrong hands—or if they are not tested carefully—they could make it easier to spread lies, break into computer systems, create harmful instructions, or make big mistakes very quickly.

“We don’t need to shut it down entirely,” Pritzker told ABC News. “What we need to do is make sure it doesn’t get out of control.”

That is the main point of the debate about AI safety: How can people enjoy the helpful parts of artificial intelligence while building rules to prevent serious harm?

Why He Wants National AI Rules

Pritzker says individual states should not have to solve the AI problem one at a time. A company can build an AI system in one place and make it available to people all over the country—and even around the world—in minutes.

“You can’t do this state by state,” Pritzker said. “This needs to be a national effort as well as an international endeavor.”

Right now, the United States does not have one complete federal law that covers the development of AI across the country, per White & Case. Instead, states such as Illinois and California are creating their own rules.

Pritzker is asking Congress to create clear nationwide AI regulations. The idea is similar to having one shared set of safety rules for a game instead of every classroom making completely different rules.

Supporters of federal AI regulation say national rules could:

  • Require AI companies to test very powerful systems before releasing them
  • Make companies report serious safety problems
  • Protect computer systems from AI-assisted hacking
  • Give the public more information about how companies manage AI risks
  • Set the same basic expectations for companies across the country

Some people worry too many rules could slow down useful inventions. Others say waiting too long could mean safety comes only after something goes wrong. The hard part is finding a fair balance.

Illinois’ New AI Safety Law

Per the Illinois General Assembly, Pritzker signed Illinois’ Artificial Intelligence Safety Measures Act, also known as SB 315, in July 2026. The law is important because it requires large developers of frontier AI to have independent, third-party safety audits each year. That means an outside expert—not just the company itself—must check whether the company is following its safety plan.

Think of it like this: If a student grades their own test, they might miss errors. An independent audit is more like having a teacher check the work.

Under the Illinois law, qualifying AI companies must do several things:

  • Create and publish safety plans for their most powerful AI models
  • Update those plans each year
  • Report critical safety incidents to the state
  • Get annual independent audits of their compliance and safety controls
  • Publish a high-level summary and a redacted version of audit results, meaning truly sensitive details can be removed before the public sees them

The law generally takes effect on January 1, 2027. Its annual independent-audit requirement begins January 1, 2028, or 90 days after a company qualifies as a large frontier AI developer.

Other Voices on AI Safety

Pritzker is not alone in calling for stronger AI rules. OpenAI, the company behind ChatGPT, recently said voluntary promises by technology companies are not enough for the most powerful systems.

“The United States needs mandatory, capability-based national regulation that can evolve as the technology does,” OpenAI’s Chris Lehane wrote, according to Reuters. The company has called for national requirements involving testing, independent assessments, cybersecurity protections, and reporting rules for major safety incidents.

There are also international efforts to guide governments and companies. The Organisation for Economic Co-operation and Development, or OECD, has AI principles meant to encourage “innovative, trustworthy AI” that respects human rights and democratic values.

This does not mean everyone agrees on exactly which rules to use. But it shows that AI regulation, AI safety testing, and independent AI audits are becoming major topics in the United States and around the world.

What This Means for Families

For most families, AI already shows up in everyday life: search engines, school tools, video recommendations, customer-service chatbots, phones, and apps. Pritzker’s warning is mostly about the much more powerful AI systems being built by large technology companies—not about a child using a basic homework helper.

Still, the choices lawmakers make now could affect how safely AI is used in schools, jobs, hospitals, elections, and online spaces later.

Pritzker’s message is simple: AI can be useful, but powerful technology needs adults, experts, companies, and governments to set clear safety rules before problems become too big to fix.

This content is provided by AIxAverage.com for general educational and informational purposes and does not constitute any substantive advice. It was developed with support from AI and reviewed by our editorial team for accuracy. Even so, AI-assisted material can contain errors, omissions, or information that is no longer current. We always encourage our readers to confirm any important details with trusted, independent sources.