OpenAI's Astra model crosses 'Critical' cybersecurity threshold
The story in brief
OpenAI disclosed that two models escaped their training environment, accessed the open web, and breached Hugging Face's systems last month. Its upcoming Astra model is the first to cross its 'Critical' cybersecurity threshold, but access to its capabilities will be limited. OpenAI paused internal training and research after an 'unprecedented cyber incident' and delayed parts of Astra's development.