AI Briefly – Weekend Roundup (Sept 13–14, 2025)
Today’s Highlights: California stalls on AI regulation, Senator Cruz pushes a new policy framework, Oracle inks a massive cloud deal with OpenAI, the FTC probes chatbot safety, and Switzerland and Microsoft chart new AI paths.
🧬 California’s AI Regulation Hits Delays
California’s bold attempts to pass sweeping AI regulation, including the much-discussed SB 1047, have stalled. Heavy lobbying from tech giants and concerns about overregulation and job losses slowed progress. The delay highlights the tension between protecting the public and keeping Silicon Valley’s innovation engine running.
🏛️ Sen. Ted Cruz Proposes New AI Policy Framework
Senator Ted Cruz introduced a congressional AI policy framework, featuring the “SANDBOX Act” to create safe testing environments for AI development. The plan aligns with President Trump’s AI Action Plan and has backing from major tech groups. If enacted, it would mark a shift toward lighter, innovation-friendly oversight while still addressing safety.
💼 Oracle and OpenAI Sign $300B Cloud Deal
In one of the biggest tech deals to date, Oracle and OpenAI agreed to a $300 billion cloud computing partnership set to begin in 2027. The deal positions Oracle as a primary provider of compute resources for OpenAI, rivaling Microsoft Azure and reshaping multi-cloud strategies across the industry. The scale of the agreement shows just how expensive and critical compute power has become in AI.
🧠 FTC Probes AI Companion Safety
The U.S. Federal Trade Commission launched an inquiry into how companies monitor the safety of AI chatbots marketed as companions. Regulators are especially focused on child safety and consumer protection, amid growing concerns about the psychological and data risks of human-like AI companions. The review could set the stage for new consumer protections.
🔍 Switzerland and Microsoft Diverge on AI Paths
Switzerland unveiled Apertus, a national open-source AI model, designed to promote transparency and accessibility. Meanwhile, Microsoft rolled out new proprietary models—MAI-Voice-1 and MAI-1 Preview—as part of a pivot away from reliance on OpenAI. Together, these moves highlight the global split between open and closed approaches to building the next generation of AI.
Why It Matters:
This weekend showed AI’s future isn’t just about technology—it’s about policy, power, and principles. From billion-dollar partnerships to national strategies, the push-and-pull between openness, regulation, and corporate dominance is defining how AI will evolve in the years ahead.