Former presidential candidate Andrew Yang is urging lawmakers to build safeguards into artificial intelligence development, including a so-called "kill switch," as concerns grow over the rapid rise of autonomous AI systems and their potential impact on cybersecurity and jobs.
A Push for Guardrails
Yang argued that policymakers need to move quickly to establish rules governing the companies building the most advanced AI models. Among his proposals are liability standards that would hold frontier AI firms accountable for the harms their systems may cause, as well as mandatory waiting periods before certain powerful technologies are deployed.
The former Democratic candidate has long positioned himself as a voice on the intersection of technology and public policy, having centered his 2020 campaign on the economic disruption caused by automation. His latest comments extend that focus to the new generation of AI tools capable of acting independently.
Yang wants a way to hit the brakes before increasingly autonomous AI systems spin out of human control.
Rising Safety Concerns
The call comes amid mounting anxiety about AI agents—software that can carry out complex tasks with minimal human oversight. Critics worry these systems could be exploited for cyberattacks or accelerate job losses across a range of industries as they take on more responsibilities once handled by people.
Yang's proposals reflect a broader debate in Washington and beyond over how to regulate a fast-moving industry without stifling innovation. Supporters of stricter oversight say the stakes are too high to leave safety measures entirely to the companies developing the technology.
Key ideas floated by Yang include:
- Liability rules holding frontier AI developers responsible for damages
- Waiting periods before deploying the most capable models
- A mechanism to shut down systems that behave dangerously
Whether lawmakers will adopt any of these measures remains uncertain, but Yang's intervention adds a prominent voice to the growing chorus demanding tighter controls as AI capabilities continue to advance.