AI News/Wire summary
DeepSeek and Moonshot AI reportedly trained on OpenAI's GPT-6 Astra outputs
Key points
- OpenAI's GPT-6 Astra, its most powerful model yet, was released in early September and has drawn strong demand.
- According to people with direct knowledge, DeepSeek and Moonshot AI leaned heavily on Astra outputs for their latest models.
- Chinese AI labs are reportedly using Astra to pull out its capabilities, a practice known as distillation.
Why it matters for marketers: Teams using Chinese AI models may need to check training provenance, since reported reliance on Astra outputs raises questions about those models' origins.
Summary by the Z Media news desk, from the outlets' reporting. Read the full stories at the source.
Read the original
The Information: OpenAI’s Astra Model Becomes Top Distillation Target for China’s AI Labs
More coverage
Related on Z Media
- Mistral Large 4 released on Hugging Face with open weightsvia Hugging Face, 6 days ago
- Nvidia releases open Nemotron 3.5 Transcribe model for 40 languagesvia TechCrunch, Oct 28
- Anthropic warns it may soon hit its own recursive self-improvement thresholdvia TechCrunch, yesterday
- Google's Suncatcher space data center prototype is online and communicatingvia TechCrunch, 4 days ago
- OpenAI restructures Model Spec around five core valuesvia OpenAI, 5 days ago
- Synopsys and Cadence shares fall 4% on Anthropic chip design reportvia The Information, 4 hr ago