OpenAI begins rolling out Astra model after warning of its advanced cyber capabilities
- On Thursday, OpenAI unveiled GPT-6 Astra, featuring advanced computer use and coding capabilities, with a phased rollout beginning through the Daybreak program and expanding to ChatGPT and API users in coming days.
- Following a July security incident at Hugging Face, OpenAI implemented stronger safeguards and designated Astra as its first model crossing the 'Critical' cybersecurity threshold under the Preparedness Framework.
- Astra achieved a 100% score on ExploitBench and 99.9% on ARC-AGI-3, marking significant performance gains over GPT-5.6 Sol and demonstrating advanced technical reasoning across cybersecurity and abstract problem-solving.
- Responding to Congressional scrutiny following the breach, OpenAI told U.S. House Democrats this week it is developing 'automated shutdown capabilities' to mitigate risks from agentic AI operating autonomously.
- OpenAI President Greg Brockman hailed the 'AGI era,' though critics caution the term remains poorly defined while the company competes fiercely against Anthropic for enterprise market share.
30 Articles
30 Articles
Ranked at the highest level of risk in cybersecurity, this new d的IA model is equipped with a surveillance device that can interrupt an unauthorized task. A reinforced prudence after several hacks conducted spontaneously this summer by tools from OpenAI and its competitorsOpenAI launched Thursday GPT-6 Astra, its first artificial intelligence (AI) model capable of hacking alone highly protected computer systems, now disseminated to the public wit…
Ranked by the company at the highest level of cybersecurity risk, GPT-6 Astra will be available "in the next few days" for ChatGPT paid subscribers, the conversational agent used by a billion people each week, according to OpenAI.
OpenAI rolls out GPT-6 Astra, and Greg Brockman says AGI has arrived
OpenAI’s president says artificial general intelligence has arrived. Greg Brockman told reporters that the company’s new Astra model qualifies, the Washington Post reported. OpenAI defines AGI as systems that are generally smarter than humans. “I do leave it up to the reader to decide for themselves if this qualifies for them“, Brockman said. “I think […] This story continues at The Next Web
Coverage Details
Bias Distribution
- 50% of the sources are Center
![[your]NEWS](/_next/image?url=https%3A%2F%2Fgroundnews.b-cdn.net%2Finterests%2Ffb6dc495f74049f513563c33352175eaa0ecd509.jpg%3Fwidth%3D60&w=128&q=75)
Factuality
To view factuality data please Upgrade to Premium
