The most sweeping proposal came from Sen. Bernie Sanders of Vermont, whose Ban Artificial Superintelligence Act would order a complete halt to the development of AI systems capable of surpassing human intelligence, according to a Thursday (Sept. 3) press release. The bill would also impose a temporary pause on the development and deployment of other advanced AI systems until a new federal regulator establishes safety guardrails.
Sanders’ proposal represents one of Congress’ most aggressive attempts to address the potentially catastrophic risks posed by AI. It would establish a cabinet-level agency charged with protecting the public from AI risks and hazards, supported by an Artificial Intelligence Advisory Board composed of experts providing independent scientific and technical advice, the release aid.
“The leaders of the major AI companies publicly acknowledge that they do not fully understand the technology and that it is escaping their control,” Sanders said in the release. “…The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs.”
One of those Big Tech oligarchs echoed Sanders’ concern. In a Sunday (Sept. 6) blog post titled “An Alien Mind,” OpenAI Chief Scientist Jakub Pachocki called for “extreme caution” over AI’s runaway progress and warned more intervention may be needed to ensure “humans remain in control of the future.”
“I am concerned no one is prepared for the consequences of a continued rapid rise in machine intelligence,” he wrote in the post.
Pachocki’s post came a few days after the ChatGPT-maker released its latest model, GPT-6 Astra, which it called its most powerful product ever.
Sanders’ legislation would back its restrictions with unusually severe penalties. A company continuing to develop prohibited systems could face the “corporate death penalty,” under which the government terminates its legal existence. Individual violators could receive prison sentences of up to 20 years, the Thursday press release said.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
The measure would move federal policy beyond the risk assessments, voluntary commitments and transparency requirements that have characterized much of the AI debate. It would instead treat the development of uncontrollable superintelligence as an activity too dangerous to permit.
In a similar vein, Reps. George Whitesides of California and Pat Harrigan of North Carolina introduced legislation directing the National Institute of Standards and Technology to monitor how advanced AI is being used to conduct further AI research and development. The measure targets the emerging risk of machines helping develop more capable machines, including systems that can autonomously facilitate scientific breakthroughs.
A more targeted bipartisan proposal would limit AI’s authority over health insurance coverage. The Doctors Not AI Act, introduced by Rep. Greg Landsman of Ohio, would prevent an automated system from having the final say on whether an insurer approves or denies a medical claim, according to a Sept. 1 press release.
AI could still play a supporting role, but denials based on medical necessity would have to be made by a licensed healthcare professional, the release said. Insurers would also have to disclose the use of AI in claims processing to patients and administrative personnel, while the bill includes protections intended to preserve access to mental healthcare.
“When a computer system is making decisions about healthcare, people will get hurt,” Landsman said in the release, warning that automated denials could leave patients with large bills, deteriorating health or life-threatening consequences.
The proposal was co-introduced by Reps. Buddy Carter of Georgia, Kim Schrier of Washington and Tom Barrett of Michigan, giving it bipartisan backing. It reflects a growing legislative focus on human accountability in high-stakes automated decisions rather than an outright prohibition on the technology.
Other bills introduced during the week focused on government AI oversight. Rep. Sara Jacobs of California proposed the Defense AI Reliability and Reporting Act, which would create a Pentagon-wide program and repository for reporting AI incidents and vulnerabilities. Jacobs said in an Aug. 31 press release that AI failures at the Defense Department could have deadly consequences or even draw the United States into a war.
Finally, Rep. Tony Wied of Wisconsin introduced a bill directing the Small Business Administration to work with the Cybersecurity and Infrastructure Security Agency on cybersecurity guidance. It would also require coordination with the Pentagon to help small federal contractors understand Cybersecurity Maturity Model Certification requirements.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
Add as Preferred Source
