---
title: "DeepSeek and Moonshot AI trained new models on GPT-6 Astra responses: The Information and Reuters reveal details"
description: "The Information and Reuters reported that DeepSeek and Moonshot AI applied GPT-6 Astra distillation when training Kimi K3.1 Pro and DeepSeek V4.1-Flash. The companies have not commented on the data."
date: 2026-10-31T17:46:07.544Z
lang: en
url: https://xab.info/en/posts/deepseek-moonshot-gpt6-astra-distillation
tags: [deepseek, moonshot-ai, openai, gpt-6-astra, ai-distillation, artificial-intelligence, kimi-k3]
author: "Alexander Buran"
publisher: "XAB.info"
---

# DeepSeek and Moonshot AI trained new models on GPT-6 Astra responses: The Information and Reuters reveal details

![Artificial intelligence interface with laptops and digital neural networks in blue glow](https://xab.info/media/2026/10/31/deepseek-moonshot-gpt6-astra-distillation/deepseek-moonshot-gpt6-astra-distillation-1.webp)

## Key Points

- The Information and Reuters reported that DeepSeek and Moonshot AI used GPT-6 Astra distillation to train Kimi K3.1 Pro and DeepSeek V4.1-Flash 1017
- OpenAI stated as early as September about a data collection campaign linked to Moonshot AI and blocked over 15,000 accounts
- Anthropic reported a 90–99% reduction in data collected via distillation after implementing new protection measures
- Representatives of OpenAI, DeepSeek, and Moonshot AI did not respond to Reuters requests for comment

Chinese artificial intelligence developers DeepSeek and Moonshot AI actively used outputs from OpenAI's GPT-6 Astra model to train their newest systems. This was reported on October 31 by The Information and Reuters, citing sources familiar with the internal processes of both companies. This concerns the practice of distillation—a method where responses from a more powerful model are used as training material for another system, allowing it to replicate certain AI capabilities without comparable computing costs.

### The distillation method and its economic logic

Distillation in the context of generative AI involves sending a large number of prompts to a stronger model, collecting its responses, and subsequently training a less resource-intensive system on that data. According to sources at The Information, this exact approach allowed DeepSeek and Moonshot AI to accelerate the development of new models without investing amounts comparable to OpenAI's expenses for creating GPT-6 Astra. For companies whose computing power lags behind the market leader, this method serves as a tool to narrow the capability gap without the need to train models from scratch on their own.

### Specific models and timeline

Reuters clarifies that GPT-6 Astra distillation results were used in training Moonshot AI's Kimi K3.1 Pro model, released last week, as well as the updated DeepSeek V4.1-Flash 1017 version, which appeared on October 17, 2026. Thus, both Chinese companies applied the same method within updates released over the past two weeks, indicating a systematic rather than isolated use of GPT-6 Astra data.

### History of the data collection campaign and OpenAI's position

The use of distillation has previously been the subject of public disputes. In a publication dated September 30, 2026, OpenAI reported that it discovered participants in a data collection campaign starting in July, which the company claimed was linked to Moonshot AI. According to ixbt.com citing the same material, OpenAI recorded activity from over 15,000 users, blocked the involved accounts by July 28, halted the campaign, and transferred the collected data to other AI model developers. OpenAI, Anthropic, and Google consistently oppose the practice of distillation, believing it allows competitors to copy their model capabilities without financial investments in research and development.

### Protection measures and lack of comments

Earlier in October 2026, Anthropic stated that new defense measures for its Claude models reduced the volume of data collected during distillation attacks by 90–99%. This indicates that leading US labs view distillation as a real threat vector to intellectual property and are actively investing in technical countermeasures. Representatives of OpenAI, DeepSeek, and Moonshot AI did not respond to Reuters requests for comment on the October 31 publications by The Information and Reuters. Neither of the Chinese companies denied or confirmed the use of GPT-6 Astra data in training their models.

## Fact-Check & Sources

- [DeepSeek и Moonshot AI использовали GPT-6 Astra для обучения своих ИИ-моделей window-new](https://3dnews.ru/1150637/deepseek-i-moonshot-ai-ispolzovali-gpt6-astra-dlya-obucheniya-svoih-iimodeley) - Primary source. Confirms the fact of GPT-6 Astra distillation for Kimi K3.1 Pro and DeepSeek V4.1-Flash 1017, citing The Information and Reuters.
- [DeepSeek и Moonshot обучали новые модели на ответах GPT-6 Astra, сообщают СМИ](https://www.ixbt.com/news/2026/10/31/deepseek-i-moonshot-obucali-novye-modeli-na-otvetax-gpt-6-astra-soobshhaiut-smi.html) - Confirms main facts. Adds detail: OpenAI recorded 15,000+ users, blocked them by July 28, transferred data to other developers.
- [DeepSeek и Moonshot AI использовали GPT-6 Astra для обучения своих ИИ-моделей](https://runet.news/news/85992) - Fully aligns with the primary source on all key facts and dates.
- [DeepSeek и Moonshot AI обучили модели на ответах GPT-6 Astra](https://tehnowar.ru/587732-deepseek_i_moonshot_ai_obuchili_modeli_na_otvetah_gpt_6_astra.html) - Confirms the main facts of the publication.

## FAQ

### What is distillation in the context of AI?

Distillation is a training method where responses from a more powerful model are used as training material for another system. A developer sends numerous prompts to a strong model, collects the responses, and trains a less resource-intensive system on them, allowing it to reproduce some capabilities without building them from scratch.

### Which specific models were affected?

According to The Information and Reuters, GPT-6 Astra distillation results were used to train Moonshot AI's Kimi K3.1 Pro (released last week) and the updated DeepSeek V4.1-Flash 1017 (released October 17, 2026).

### Did DeepSeek and Moonshot AI comment on these reports?

No. Representatives of OpenAI, DeepSeek, and Moonshot AI did not respond to Reuters requests for comment on the October 31, 2026 publications.

### What did OpenAI do regarding the data collection campaign?

In a publication dated September 30, 2026, OpenAI reported discovering a data collection campaign that began in July, allegedly linked to Moonshot AI. By July 28, the involved accounts were blocked, the campaign was stopped, and the collected information was shared with other AI model developers.