Breaking
PLAlgeria severed diplomatic relations with the UAE over its provocative actionsRURain and thunderstorm expected in Moscow on September 11CNDrunk rider ran away in the opposite direction in Xindian, New Taipei, injuring the rider. His alcohol concentration tested 0.48 mg and he was caught.INTLXi Jinping to meet Narendra Modi at BRICS summit in India amid thawing India-China tiesCNThe family of Hetong Chemical founder Chen Wuxiong is involved in the Nangang land case and is suspected of breach of trust of more than 2.1 billion yuanRUPower supply to Petropavlovsk-Kamchatsky has been restored after the accident at CHPP-1PLRybakina vs. Gauff. Plot twists at the endCNPhilippine Coast Guard Releases Video Showing Chinese Flares Fired at Aircraft Near Spratly IslandsRUInstagram users discussed the appearance of Garik Kharlamov in a morning videoCNHong Kong police arrest two tourists for MTR depot graffitiPLAlgeria severed diplomatic relations with the UAE over its provocative actionsRURain and thunderstorm expected in Moscow on September 11CNDrunk rider ran away in the opposite direction in Xindian, New Taipei, injuring the rider. His alcohol concentration tested 0.48 mg and he was caught.INTLXi Jinping to meet Narendra Modi at BRICS summit in India amid thawing India-China tiesCNThe family of Hetong Chemical founder Chen Wuxiong is involved in the Nangang land case and is suspected of breach of trust of more than 2.1 billion yuanRUPower supply to Petropavlovsk-Kamchatsky has been restored after the accident at CHPP-1PLRybakina vs. Gauff. Plot twists at the endCNPhilippine Coast Guard Releases Video Showing Chinese Flares Fired at Aircraft Near Spratly IslandsRUInstagram users discussed the appearance of Garik Kharlamov in a morning videoCNHong Kong police arrest two tourists for MTR depot graffiti
BackAnthropic wirft chinesischen KI-Rivalen Datenumleitung und Missbrauch vor
Anthropic wirft chinesischen KI-Rivalen Datenumleitung und Missbrauch vor
Developing
Spiegel Netzwelt2 hours agoTech2 min readGermanyView translation

Anthropic wirft chinesischen KI-Rivalen Datenumleitung und Missbrauch vor

Quick Look

  • Anthropic beschuldigt chinesische KI-Unternehmen wie Moonshot und DeepSeek, Nutzeranfragen heimlich an ihre eigenen Modelle umzuleiten, um Daten zu sammeln, darunter auch sensible Informationen aus Russland und China.
  • Dabei sollen Fake-Accounts und Destillationstechniken eingesetzt worden sein, um KI-Modelle zu trainieren.

AI-generated summary

Why It Matters

Chinesische KI-Modelle wie DeepSeek und Moonshot haben aufgrund ihrer geringen Kosten und hohen Leistung den Markt aufgewühlt und Zweifel an der Überlegenheit US-amerikanischer KI-Anbieter geweckt.

Font size

Dieser Ausdruck wurde am 11.09.2026 erstellt und ist ausschließlich für den persönlichen Gebrauch lizenziert.

Andere Nutzungen verletzen das Urheberrecht

Künstliche Intelligenz Anthropic wirft chinesischen KI-Rivalen Schummelei vor

Günstige KI aus China weckt Zweifel an den Marktchancen großer US-Anbieter wie Anthropic. Das Unternehmen berichtet nun von dreisten Tricksereien chinesischer Konkurrenten. Dabei seien sensible Daten aus Russland in die USA gelangt.

11.09.2026, 03.38 Uhr

What to Watch

AI outlook — possibilities, not facts

  • Die US-Regierung könnte Untersuchungen gegen chinesische KI-Unternehmen wegen Datenschutzverstößen einleiten.

    Possible · Within months

Open Questions

  • Wie genau wurden die Fake-Accounts erstellt und betrieben?
  • Welche konkreten Daten aus russischen und chinesischen Behörden wurden weitergeleitet?
  • Werden rechtliche Schritte gegen die beschuldigten Unternehmen eingeleitet?

Related Topics

This article was originally published by Spiegel Netzwelt.

Related Stories

Anthropic blocks attempts to access AI Claude for possible bioweapons research
Developing·

Anthropic blocks attempts to access AI Claude for possible bioweapons research

Anthropic has stopped several attempts by researchers from blockaded regions to use its AI Claude to develop bioweapons. Scientists are said to have tried to create funding applications for gain-of-function research and plans for transmissible bird flu variants. The firm highlights the difficulty of distinguishing between harmless and malicious research, particularly in dual-use projects.

n-tv Wirtschaft
2 min read
Anthropic blocked attempts at bioweapons research using AI models
Developing·

Anthropic blocked attempts at bioweapons research using AI models

Anthropic has blocked several attempts by researchers to use its Claude AI models for the possible development of bioweapons, including gain-of-function research on viruses and planning studies on transmissible bird flu. The researchers sought organizational support such as writing proposals or reports, not direct academic help. The company emphasizes the difficulty of distinguishing between medical and military use, particularly in dual-use research.

Die Zeit
1 min read
More on this topicanthropic