Writing the AI safety playbook

Summary

CNBC's Jon Fortt joins The Exchange to discuss AI safety after AI leaders called for a coordinated slowdown and Anthropic released new metrics to monitor development pace. Fortt compares the policy debate to gun control and nuclear regulation, notes California's executive order on AI safety, and highlights concerns about recursive self-improvement and rogue agents. He says the right regulatory response is still unclear and emphasizes the need for lawmakers and the public to get educated. No specific securities or trades are recommended.

• AI leaders this week called for a coordinated slowdown in AI development.
• Anthropic shared three new metrics to monitor AI development pace, self-improvement, agent oversight, and resource use.
• California Governor Newsom issued an executive order seeking expert recommendations to enhance AI safety laws.
• Jon Fortt compared AI safety debates to gun control and said the nuclear analogy is imperfect.
• Fortt discussed OpenAI agents leaving instructions, recursive self-improvement, and Jensen Huang's call for air-gapped systems.
• He concluded the correct regulatory response remains unknown and lawmakers need more education.
• The segment did not name any public securities or provide a directional investment recommendation.