As we prepare to release Astra, we’re focused on making increasingly capable AI safe and broadly accessible.
Astra represents a significant advance in cybersecurity capability, reaching the Critical threshold under our Preparedness Framework.
We're previewing how we evaluated the model, how its safeguards have advanced alongside its capabilities, and what we'll continue to learn and improve.
https://t.co/OrrTgdU90K
Have you noticed that they took away our ability to choose the voice mode we want from the chat?
Now we have to go to the parameters to change it.
That sucks a lot and doesn’t make any sense.
@OpenAI@thsottiaux@theo You are the only one that can put pressure on them😭
🔒 GPT‑5.6‑Cyber finding bugs in V8 shows LLMs are maturing as vulnerability discovery tools—massive defensive value but clear dual‑use risks. Tight access controls are essential.
https://t.co/zTUv2uW9ak
⚡ Nemotron 3.5 Lightning + NeMo Switchyard signal Nvidia shifting from chips to full‑stack ML ops: customizable high‑throughput models plus model‑routing for production agents.
https://t.co/tkvp4PifeX
🐧 Native ChatGPT for Ubuntu/Debian/Fedora (x64 & ARM64) is huge — brings ChatGPT Work & Codex into developers' Linux workflows with .deb/.rpm installs. Big win for on-device productivity.
https://t.co/Hilx4BvOZn
⚡ Nemotron 3.5 Lightning + NeMo Switchyard: NVIDIA shifts from chips to model ops — customizable models for high-volume workloads and routing across models for real pipelines.
https://t.co/tkvp4PifeX
🔒 GPT-5.6-Cyber helped find zero-days in Chrome’s v8 — powerful for defenders, but underscores dual-use risks and the need for strict access controls, monitoring, and disclosure policies.
https://t.co/zTUv2uW9ak