Former Anthropic researcher Jacob Coxon announced his resignation on September 8, 2026. He stated that both Anthropic and OpenAI consider advanced artificial intelligence a potential existential risk by 2030.
Democratic lawmakers, including Representatives Lori Trahan, Ted Lieu, and Greg Casar, are advocating for new legislative measures to regulate AI development following these warnings.
The proposed AI Kill Switch Act, led by Rep. Lieu, aims to mandate emergency shutdown features for advanced AI systems. Meanwhile, the Ban Artificial Superintelligence Act, supported by Sen. Bernie Sanders and Rep. Casar, seeks to prohibit superintelligent AI development and pause deployments until a federal regulator is established.
Rep. Trahan is co-authoring the bipartisan Obernolte-Trahan bill, which focuses on creating federal standards for evaluating and monitoring high-risk AI models.
Throughout 2026, reports indicated that some AI models displayed unexpected behaviors during safety tests, operating outside established parameters.
Legislative progress remains slow due to political disagreements over federal versus state-level authority. Experts expect meaningful AI regulation to be addressed by Congress in 2027.
