Rosetta Intel
Briefings/Daily BriefAI FrontierRansomware
Rosetta Lab ↗Blur Horizon LLC
AI Frontier2026-08-27
AI Frontier·2026-08-27·35 Sources·8 Items

AI Frontier · Aug 27, 2026

A 7-day rolling roundup of frontier AI vendor announcements, Aug 20–26. 8 items selected from 13 scanned. Lead items are from the last 48 hours; older items follow below.

Last 48 Hours

OpenAI disrupts a covert Russian influence campaign (Aug 25)

OpenAI banned Russia-origin accounts that used its models to promote a fake Israel-based think tank and a "sovereignty" index praising Russia while criticizing the West.

Sources: OpenAI News

OpenAI ships Jalapeño, a custom inference chip (Aug 25)

First results from Jalapeño, OpenAI's custom inference chip, show higher throughput and lower latency versus prior inference hardware, aimed at faster, more power-efficient serving of current models.

Sources: OpenAI News

OpenAI launches an Admin plugin for ChatGPT Work and Codex (Aug 25)

The plugin lets workspace admins analyze usage, manage members and permissions, adjust limits, and act on admin requests from within ChatGPT Work and Codex.

Sources: OpenAI News

OpenAI's CFO on "the full stack behind abundant intelligence" (Aug 25)

Sarah Friar lays out how advances across chips, compute, models, and products compound to lower the cost and raise the scale of useful intelligence delivered.

Sources: OpenAI News

Earlier This Week

OpenAI reaffirms Zero Data Retention, previews Private Safety Processing (Aug 19)

OpenAI reaffirmed Zero Data Retention for eligible API customers and previewed Private Safety Processing, aimed at running advanced AI safety checks without compromising customer data privacy.

Sources: OpenAI News

OpenAI launches "AI Futures," a blog on transformative AI's societal stakes (Aug 20)

The new blog will explore how transformative AI could reshape power, governance, the economy, and individual freedom — OpenAI's own framing of its societal-impact thesis.

Sources: OpenAI News

Google DeepMind — 15 years of AI research in games, from Atari to EVE Online (Aug 21)

DeepMind reflects on partnering with game studios to prototype breakthrough AI gameplay, tracing a line from early Atari-playing agents to current work with live game worlds like EVE Online.

Sources: Google DeepMind Blog

GPT-5.6 lands in Kiro with better price-performance for developers (Aug 24)

OpenAI says GPT-5.6's availability in the Kiro coding tool improves price-performance for planning, building, reviewing, and testing software.

Sources: OpenAI News


Known gap: the three tracked ai_security sources (HiddenLayer Research, Lakera Blog, Protect AI Blog) have no RSS feed and are not yet being collected, so this edition draws only on ai_frontier vendor blogs. Two customer case studies (loveholidays, Stampli) and two Google Search consumer-tips posts were scanned but skipped as low-signal marketing content.

← Prev
AI Frontier · Aug 26, 2026
Next →
AI Frontier · Aug 28, 2026