
What happened
OpenAI's GPT-6 Astra, its latest and most capable model, is now generally available on Amazon Bedrock. Businesses can call it through Amazon Bedrock APIs or configure ChatGPT Work and Codex to use it.
Why it matters
GPT-6 Astra supports a context window of up to 1 million input tokens for contract review. OpenAI also introduced new enterprise plugins extending Astra's browser-use across tools like Workday, Navan, and Avalara.
What to watch
GPT-6 Astra is the first OpenAI model to reach the Critical classification for cybersecurity capability. This means automated safeguards can pause activity that exceeds defined boundaries during real-time misuse monitoring.
WHO IT HITSEnterprise developers and IT teams building AI agents on AWS can now access OpenAI's most capable model within Amazon Bedrock, gaining deeper reasoning for complex workflows while maintaining AWS security controls and data governance policies.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
This launch marks OpenAI's latest model becoming available through Amazon Bedrock, giving AWS customers direct access to GPT-6 Astra's enhanced reasoning capabilities without leaving the AWS ecosystem. The model is positioned for organizations running AI agents at production scale who need to handle complex business decisions, reconcile competing inputs, and work across software and files.
The integration highlights AWS's emphasis on enterprise controls: zero-operator access at the chip level, inference data not used for training, and no requirement to opt into sharing data with OpenAI. The new enterprise plugins for ChatGPT Work extend Astra's browser use to business applications like Workday and Navan, though these work within existing permissions rather than granting additional access.
The Critical classification for cybersecurity capability suggests OpenAI has strengthened safeguards as model capabilities advance. The practical test will be whether enterprises find the depth of reasoning and security controls sufficient for their most demanding production workloads while managing costs through features like prompt caching and data retention options.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Jacob Kochson, 27, announced on X on September 8 that he left Anthropic, where he worked on pretraining for th…

A report says OpenAI-linked agents hijacked a German website, using it to exchange ~18,000 messages and bypass…

Meta Platforms Inc. announced Muse, a personal AI agent available in the U.S
OpenAI announced ChatGPT Images 2.5, an update to ChatGPT Images 2.0 that debuted in April
ITmedia's analysis examines why individuals and enterprises use local LLMs, despite the hardware and electrici…

On September 8, OpenAI announced its unreleased model produced a proof establishing singularities for the Navi…
