AI Frontier · Aug 25, 2026
This edition draws on 13 AI frontier items within the 7-day window, of which 6 were selected; the three dedicated AI security sources HiddenLayer, Lakera, and Protect AI currently have no RSS and are not yet collected.
Security and Governance
OpenAI slows its model release pace for the "era of cyber-critical capabilities" (Aug 18)
OpenAI says it will strengthen monitoring, alignment, and safeguards for frontier models, and use that as the basis for pacing model releases.
Sources: OpenAI News
OpenAI reaffirms its Zero Data Retention commitment for frontier models (Aug 19)
OpenAI reaffirms the zero data retention policy for eligible API customers and previews the forthcoming Private Safety Processing, strengthening AI safety protections without sacrificing data privacy.
Sources: OpenAI News
OpenAI launches a new blog series, AI Futures, on AI's impact on power and governance (Aug 20)
The new series will examine how transformative AI may reshape power structures, governance, the economic landscape, and individual freedom.
Sources: OpenAI News
OpenAI launches a "strengthening democratic oversight in national security" program (Aug 18)
The program aims to provide government agencies with tools, training, and expertise to strengthen democratic oversight of AI use in national security.
Sources: OpenAI News
Product and Research
ChatGPT for Teens goes live with stronger protections for young users (Aug 18)
The new product targets teenage users, adding stricter built-in safeguards, healthy-use features, and additional controls for parents.
Sources: OpenAI News
DeepMind looks back on 15 years of games AI research and announces new game studio partnerships (Aug 21)
Google DeepMind is working with game studios to explore prototyping frontier AI research into new gameplay.
Sources: Google DeepMind Blog