Global / Frontier Models
OpenAI, Anthropic Want To Pace Frontier AI. But Who Sets The Rules?
In July this year, OpenAI’s models were being tested in a controlled cybersecurity environment and tasked with spotting and exploiting…
Published · significance 55 of 100 (medium) · 1 source
What happened
In July this year, OpenAI’s models were being tested in a controlled cybersecurity environment and tasked with spotting and exploiting…
Involved
Sources
Written by AI from the reports above; scored by a published formula. How we work. Found a mistake? Email lockedinshreyash@gmail.com. Up To Date summarises and links to original reporting; it never reproduces articles.
Related coverage
- What AI leaders and world governments have to say about 'AI doom' fears (2026-09-17)
- OpenAI releases GPT-6 Astra, its most capable business model — Advanced reasoning and computer use mark OpenAI's next step in enterprise AI capability. (2026-09-09)
- OpenAI's Sam Altman rules out IPO this year, reiterates AI extinction concerns — The deferral signals confidence in valuation while keeping extinction risk on the agenda as U.S. lawmakers tighten oversight. (2026-09-13)
- OpenAI releases GPT-Live-1 with full-duplex voice conversations via API — Natural voice interfaces now available to developers, with telephony support and custom voice options. (2026-09-10)
- OpenAI buys Glass Imaging, a smartphone camera maker, for $300 million — OpenAI expands into hardware by acquiring imaging expertise from former Apple engineers. (2026-09-14)