About this episode

The episode discusses OpenAI's pause in training and a cyber attack on Taiwan's nuclear safety agency.

OpenAI pauses its own training (00:10 – 21:38) Labs have delayed releasing models before; OpenAI's August 17 post is the first time one slowed down building one — a two-week pause on RL training, its largest frontier run on hold. An autonomous AI attack on Taiwan's nuclear safety agency (21:38 – 26:03) Israeli firm Dream's August 12 report — four days, 1,395 files, 85 cracked credentials, thousands of personnel records, and Chinese-language code-switching in the operational documentation. 5...

People in this episode

Host: Gregory C. Allen

Topics covered

• AI training
• cybersecurity
• data privacy
• international relations
• technology policy

Keywords

• OpenAI
• AI training
• Taiwan
• cyber attack
• data security
• Dream
• nuclear safety

Mentioned in this episode

Organizations: OpenAI, Dream, Israeli

Places: Taiwan

More episodes of The AI Power Podcast

• GPT-6 Astra - WOW, OpenAI = pro regulation, Trump admin warms to AI safety · September 11, 2026 · 1h 3m
• OpenAI hack changes state AI regulation debate + Chinese AI running on Chinese chips? · August 28, 2026 · 48 min
• AI's Impact on Education with Casey Cuny - 2024 Cal Teacher of the Year · August 26, 2026 · 1h 2m
• Is China Entering the Global Memory Supply Chain? · August 11, 2026 · 36 min
• MORE AI Cyber and Bio, Chinese Models, and a U.S. Regulatory Scramble · August 7, 2026 · 1h 16m
• China's homegrown DUV lithography breakthrough and the debate over banning China's open-source AI · July 30, 2026 · 57 min

Explore listener stats, chart rankings, contacts and more on the The AI Power Podcast podcast page.