OpenAI accuses Moonshot AI of data extraction efforts
Summary
OpenAI has accused its Chinese competitor Moonshot AI of orchestrating a large-scale data extraction effort aimed at its GPT artificial intelligence systems, which could enable the reproduction of the reasoning and capabilities of OpenAI’s most advanced models. This accusation comes in light of warnings issued in September 2026 by U.S. agencies, including CISA, NSA, and FBI, regarding China-based AI companies engaging in extensive knowledge distillation campaigns targeting U.S. frontier models. Additionally, other industry players like Anthropic have raised similar concerns about Moonshot AI's practices.
Analysis
OpenAI: OpenAI is a leading artificial intelligence research and deployment company known for developing advanced generative models such as the GPT series. In this news, it publicly disclosed and disrupted what it described as a coordinated campaign linked to Moonshot AI users attempting to extract protected reasoning data from its models through adversarial distillation techniques. Moonshot AI: Moonshot AI is a Chinese artificial intelligence company that has developed the Kimi chatbot and other competitive models. It is directly named in the news as the entity whose associated users played a significant role in efforts to extract data from OpenAI's GPT systems to replicate advanced capabilities. Industry Scrutiny: Anthropic has previously raised similar concerns about Moonshot AI routing requests to its models to support training of Kimi. Security Advisory: In early September 2026, US agencies including CISA, NSA, and FBI issued warnings about China-based AI companies engaging in large-scale knowledge distillation campaigns targeting US frontier models. International Dialogue: US and Chinese leaders met in September 2026 and agreed to establish a formal dialogue mechanism on AI safety and related risks.
Categories
aitechai_agentsmachine_learning
Related sources
- https://techcrunch.com/2026/09/25/for-months-openais-agent-swarms-have-been-attacking-online-databases-to-find-obscure-facts/
- https://www.reuters.com/legal/litigation/openai-says-ai-models-accessed-australian-government-systems-without-2026-09-29/
- https://www.recodechinaai.com/p/us-and-china-open-an-ai-dialogue
- https://www.scientificamerican.com/article/chinas-open-ai-models-are-testing-americas-approach-to-ai-safety/
- https://www.nytimes.com/2026/09/25/technology/openais-ai-us-government-websites.html/
- https://techcrunch.com/2026/09/17/microsoft-exec-called-ai-scraping-the-largest-theft-of-labor-in-human-history-new-unredacted-filings-reveal/
- https://www.scmp.com/topics/chatgpt-and-other-generative-ais
- https://www.theverge.com/openai
- https://www.cnbc.com/2026/09/26/china-ai-global-adoption.html
- https://www.ft.com/stream/e3402603-d253-4aa1-ac4d-fc9bdbf4ccb8
- https://superintelligencenews.com/applications/openai-agents-probe-databases-transluce-finds/
- https://www.theverge.com/ai-artificial-intelligence/1002854/openai-claims-moonshot-extracted-its-data-to-train-ai-models
- https://www.cisa.gov/news-events/news/cisa-nsa-and-fbi-warn-china-based-ai-companies-targeting-us-ai-models-industrial-scale-knowledge
- https://m.economictimes.com/tech/artificial-intelligence/china-and-the-us-are-competing-for-ai-dominance-but-have-shared-concerns-over-safety/articleshow/134380222.cms
- https://www.bloomberg.com/news/articles/2026-09-30/openai-blames-moonshot-for-mass-data-extraction-on-its-ai-models
- https://financialpost.com/technology/tech-news/openai-blames-moonshot-for-mass-data-extraction
- https://www.digitaljournal.com/article/ai-warning-shots-focus-beijing-on-national-security-risks/
- https://www.benzinga.com/markets/private-markets/26/09/62095522/openai-says-moonshot-ai-linked-users-tried-to-extract-its-models-secret-reasoning
- https://www.hudson.org/technology/ai-revolution-could-become-national-security-disaster-if-we-arent-careful-alexandra-seymour
- https://www.cnbc.com/2026/10/01/openai-chinas-moonshot-ai-kimi.html