Former U.S. President Barack Obama warned at a closed-door fundraising event, according to The New York Times, that artificial intelligence (AI) can be dangerous if not properly managed. He urged Democrats to develop a clear framework as soon as possible and to prioritize addressing AI’s safety and economic risks.
Obama also suggested that candidates running in the 2028 presidential election make AI a central campaign issue and present "very clear plans" to handle concerns related to the technology. Within the Democratic Party, where leadership is contested, many look to Obama for guidance; his comments carry weight ahead of the November midterm elections.
Growing concern within the industry
Concerns about risks associated with AI development have increased pressure on legislators and corporate leaders to do more. This week Jacob Coxon, a researcher at Anthropic, resigned, stating that "those building artificial intelligence seriously believe it could kill us all by the end of the decade." Reuters reports that Anthropic’s chief executive has also signaled a pullback.
Dario Amodei called on companies active in AI to slow the pace of their model development.
Amodei’s proposals: a three-step framework and independent evaluators
Dario Amodei outlined a three-step framework intended to regulate development speed and allow more time to address risks. The proposal includes bringing in independent evaluators who would have employee-level access to inspect safety practices.
Amodei also advocated greater cooperation among AI companies to set safety standards and limit uncontrolled development.
Industry leaders express support
Other industry leaders responded to the proposals. Both Elon Musk, head of xAI, and Sam Altman, CEO of OpenAI, indicated agreement with Amodei’s ideas. Sam Altman said, "It’s an excellent idea to use independent evaluators with employee-like access; we will do that."
These statements and internal tensions highlight that managing AI-related risks will require action not only from technology companies but also from policymakers if broader societal and economic harm is to be avoided.
