Axios Live: Americans don't have to accept AI risk, policymaker says

SAN FRANCISCO — Balancing AI innovation with safety requires stronger oversight, targeted development and greater public education about the technology, business and policy leaders said at an Axios roundtable event.

Why it matters: As AI becomes more integrated into Americans' daily lives, businesses and government operations face growing pressure to address potential harms without stifling innovation.

Axios' Nathan Bomey moderated the Oct. 6 roundtable discussion. The event was sponsored by IBM.

5 big takeaways from the conversation:

1. Accepting an intolerable level of AI risk is "a false choice," said Rep. Sam Liccardo (D-Calif.), who warned that an expectation of "tolerance for bad things" could justify regulatory inaction.
2. Americans aren't sufficiently prepared for how AI is going to influence their lives, said aiEDU founder and CEO Alex Kotran. "If we don't figure that out … we're still going to have a crisis."
3. Feeding superintelligent models "everything and anything" may heighten runaway AI concerns, said Overlook Ventures founder and managing partner Amir Kabir. Developing more specialized models that stay in their own lane, such as voice or text tools, could help limit those risks.
4. Third-party audits can help enterprises maintain accountability to themselves and customers as they adopt agentic AI, said Workday vice president and chief privacy and digital trust officer Barbara Cosgrove.
5. America should mitigate "the most catastrophic risks" so that it can focus on AI's economic and political implications, which the country is "better" equipped to manage, said ID.me CEO Blake Hall.

Content from the sponsor's remarks:

During opening remarks, IBM senior vice president and chief global affairs officer Jonathan Adashek said he was eager to discuss balancing digital control and innovation.

• "[I'm] excited to talk about this topic. I think it's an important one that a lot of people are focusing on."