Monitor this situation.

Unsubscribe anytime.

[SITUATION] · [QUIET] · [TECHNOLOGY]

8 clusters · 111 sources · 32 days · First seen about 2 months ago · Last updated 18 days ago

Allegations of AI data redirection by Chinese firms

Overview

Moonshot AI has come under scrutiny following allegations from Anthropic regarding its Kimi K3 artificial intelligence model. Anthropic claims the company used thousands of accounts to route approximately 300,000 user queries to its own Claude models, presenting the responses as Moonshot’s own work. Following these allegations, the Cyberspace Administration of China (CAC) launched an investigation into Moonshot AI and DeepSeek. A September 2026 threat intelligence report from Anthropic expanded these claims, alleging that several Chinese firms engaged in “illicit distillation.” Specifically, Anthropic alleged that DeepSeek rerouted over 12.1 million queries to Claude within a 14-day period in July to capture the model’s “thinking signature” for cheaper training. Recent details from the Anthropic report suggest the scale of Moonshot AI’s activity may be significantly higher, alleging the company routed approximately 23 million user prompts to Claude between May and July 2026. The report further claims these interactions included sensitive information, such as data from a Russian government agency and a Chinese municipal police system. In response to the report, the CAC began investigating whether sensitive Chinese military, police, and corporate data was inadvertently transferred to U.S. servers, potentially violating China’s data security laws. In early October 2026, OpenAI joined the scrutiny, announcing it had blocked a coordinated campaign aimed at extracting the protected internal reasoning processes of its GPT models, which it described as “adversarial distillation.” OpenAI traced the actors to individuals associated with Moonshot AI, noting a surge of 16,000 extraction requests in late July. On October 18, 2026, the U.S. Department of Homeland Security (DHS) ordered the removal of DeepSeek products and any third-party services utilizing its models from all U.S. civil agency IT systems. On October 24, 2026, the United States expanded these restrictions, implementing new sanctions and bans targeting Chinese AI entities.

Entities

DeepSeek · China · United States · U.S. Department of Homeland Security · Moonshot AI

Claims

What the coverage asserts, and how many sources carry each claim.

[● 40 SOURCES] The ban extends to third-party services and products that were developed using DeepSeek models.
[● 39 SOURCES] The DHS expressed concern that DeepSeek could result in the transfer of sensitive data to China.
[● 35 SOURCES] The DHS has ordered all federal civilian agencies to remove DeepSeek products and services from their IT systems.
[● 31 SOURCES] The DHS order prohibits federal civilian agencies from purchasing DeepSeek products or services.
[● 29 SOURCES] The DHS identified DeepSeek as presenting multiple threat vectors.
[● 29 SOURCES] The DHS order excludes Department of Defense and intelligence community agencies from the ban.
[● 21 SOURCES] China’s Foreign Ministry spokesperson Mao Ning rejected the ban.
[● 20 SOURCES] Agencies may apply for exemptions within 60 days of the order.
[● 17 SOURCES] The order was signed by Secretary Markwayne Mullin on July 24, 2026.
[● 12 SOURCES] The Cyberspace Administration of China (CAC) has launched an investigation into DeepSeek and Moonshot AI regarding data security.
[● 11 SOURCES] Anthropic accused seven Chinese AI companies of using illicit distillation to train their models using Claude's outputs.

Timeline

18 days ago

[TECHNOLOGY] 5 sources
United States bans DeepSeek from federal civilian agencies
The US Department of Homeland Security has banned DeepSeek and related third-party technologies from federal civilian agency systems due to data security and supply chain concerns.
18 days ago

[TECHNOLOGY] 6 sources
United States imposes sanctions and bans on Chinese AI entities
The U.S. has banned federal agencies from using DeepSeek and sanctioned MiraclePlus and Shanghai Zhisuan Technology over national security and AI technology theft concerns.
21 days ago

[TECHNOLOGY] 2 sources
DeepSeek and MiniMax receive approval for global use amid international scrutiny
DeepSeek and MiniMax have secured approval for global operations in China, even as the US, Australia, and several European nations restrict the firm due to data security and privacy concerns.
21 days ago

[TECHNOLOGY] 8 sources
US Department of Homeland Security bans DeepSeek from federal systems
The US Department of Homeland Security has ordered federal civilian agencies to remove DeepSeek AI from their systems and banned new purchases due to national security and data privacy concerns.
23 days ago

[TECHNOLOGY] 74 sources
US DHS bans DeepSeek AI from federal civilian agencies
The U.S. DHS has banned DeepSeek AI from federal civilian agencies, citing risks of sensitive data transfers to China and potential malware in the technology.
about 1 month ago

[TECHNOLOGY] 2 sources
OpenAI accuses Moonshot AI associates of attempting to extract GPT reasoning
OpenAI has blocked a campaign involving over 15,000 users attempting to extract protected internal reasoning from GPT models, tracing core actors to individuals linked to China's Moonshot AI.
about 2 months ago

[TECHNOLOGY] 2 sources
China investigates AI firms DeepSeek and Moonshot AI over data security
China's cyberspace regulator is investigating AI leaders DeepSeek and Moonshot AI over concerns that sensitive state and military data may have been transferred to US-based servers.
about 2 months ago

[TECHNOLOGY] 16 sources
DeepSeek and Moonshot AI under investigation by China over data security
China is investigating AI firms DeepSeek and Moonshot AI after Anthropic accused them of using Claude's outputs to train their own models via illicit distillation and unauthorized data routing.

Sources

ameve.eu · api.metro1.com.br · arynews.tv · astig.ph · berita.rtm.gov.my · bitcoinethereumnews.com · blocktempo.com · borders.gov.au · borncity.com · brasil247.com · breitbart.com · cafebiz.vn · canaltech.com.br · capital.fr · chip.com.tr · cn.nytimes.com · coingeek.com · compsmag.com · consumerismo.it · convergenciadigital.com.br · correiomanha.pt · crypto-times.jp · cryptopolitan.com · cyber-securite.fr · cybernoz.com · decrypt.co · digitaljournal.com · dijitaliyidir.com · donanimgunlugu.com · drweb.de · dunya.com · eloutput.com · enerzine.com · english.alsiasi.com · enter.co · epochtimes.com · expedientesinico.com · fastcompanybrasil.com · finance.technews.tw · forkast.news · forux.it · fr.businessam.be · gamemag.it · gigazine.net · greenmountain.no · ichongqing.info · infoarenales.com · information.tv5monde.com

This summary has been updated 16 times: see revision history