AI race heats up as OpenAI flags alleged model-copying campaign
OpenAI says it identified an attempt to extract protected reasoning from its AI models, linking part of the activity with China’s Moonshot AI.
TL;DR
- OpenAI identified and disrupted a campaign attempting to extract protected reasoning from its AI models.
- A core cluster of this activity was linked to Chinese startup Moonshot AI, developer of Kimi.
- The campaign involved manipulating interactions with OpenAI's models to reproduce hidden reasoning.
- OpenAI described the activity as "adversarial distillation," which could pose safety and national security risks.
- This follows similar accusations by rival Anthropic against Chinese AI developers.
- OpenAI has shared its findings with other AI developers and government channels.