China news roundup / Nachrichten 2026-09-10/11

The questions China won’t answer about the Tibet-Nepal flood disaster

"For years, analysts and observers have warned about the dangers of China’s relentless infrastructure push in Tibet, particularly its dam-building ambitions on the Brahmaputra River, known in China as the Yarlung Tsangpo. Beijing has repeatedly waved away such concerns, insisting it has conducted feasibility studies and environmental impact assessments."

--- How much you can trust Chinese feasibility studies and environmental impact assessments can be seen anywhere in the world where you have Chinese companies constructing major infrastructure projects.

"Treating the floods as a one-off tragedy would be a mistake; they may be only an initial warning of what lies ahead."

--- They surely are a warning, but in this particular case it's very hard to pin it on a Chinese infrastructure project. Just like you shouldn't automatically pin it on climate change.

Twenty-five dead after fire on cargo ship in eastern China

"Twenty-five people have been killed in a fire on a foreign-flagged cargo ship docked at a port in eastern China"

--- This is one of the few cases in which the Chinese victim count seems realistic. Then again...

"The country's leader Xi Jinping earlier called for "all-out" search and rescue effortsThe fire comes nearly a month after a shipyard explosion in the south-eastern Chinese province of Fujian killed a rescue worker and injured 12 others."

--- If it is bad enough that Emperor Xi gets involved, you have to ask yourself what's really going on.

For China, a Mock A.I. Attack on WeChat Signals a Dangerous New Era

"with A.I., even a tiny team with no government backing can mount a devastating attack on an app used by 1.4 billion people a month [...] the tool, which they named WeWorm, could hijack a WeChat user’s account, call their contacts and then spread from phone to phone without anyone ever answering a call. Calif said it took the company a little more than a week to build the bug"

--- WeChat is a super app which almost everyone in China uses. It has pretty much everything included, bank account and all. Would be quite interesting to see someone actually using AI to attack it. Could be a disaster for China.

"This discovery adds fuel to warnings from researchers that A.I. hacking capabilities are advancing faster than defenses can keep up, with tech leaders like Bill Gates warning that addressing risks from A.I. should be “the world’s top priority.”"

--- At the moment the priority seems to be creating risks.

"New risks seem to surface almost daily. Anthropic, the company behind Claude and other A.I. models, said in a report on Thursday that it had disrupted plots by scientists to use its models for research that could have helped develop biological weapons."

--- Looks like we don't have to wait for AI to become powerful enough to destroy mankind. 'Scientists' using AI as a tool might be the more probable culprits. Considering where and how COVID came to be, China is also the place where most probably the destruction of mankind will originate.

"Some Chinese scholars have suggested that American and Chinese officials could potentially discuss voluntary “negative lists” of unacceptable A.I.-enabled cyber activities or identify categories of critical infrastructure that are protected"

--- Yeah, and we could definitely trust the CCP to actually stick to such negative lists.

Chinese AI labs secretly used millions of Claude exchanges to train their models, Anthropic says

"Anthropic said operators affiliated with Alibaba used Claude outputs to help train its Qwen models, while Moonshot routed some Kimi user requests to Claude and used some of the resulting exchanges to train its own models. The Alibaba operation was the largest distillation campaign Anthropic said it has measured, involving more than 151 million exchanges with Claude between May and July. The activity peaked at nearly 3 million exchanges per day from more than 3,500 fraudulent accounts, according to the report. The company said Alibaba also used Claude for broader AI research, including reinforcement learning and model architecture. Anthropic also detailed activity involving Moonshot AI, the Beijing-based company behind the Kimi family of AI models. Moonshot silently forwarded some customer requests intended for Kimi to Claude and then displayed Claude’s responses to users, who thought they were using a Kimi model, according to the report. In one 10-day period, Moonshot relayed nearly 300,000 customer requests to Anthropic, the vast majority of which were routed to Claude Opus models. The requests were routed through a network of 5,380 accounts that Anthropic described as fraudulent, most of which appeared to be located in Singapore and Japan."

--- Innovation with Chinese characteristics. "Oh look, the results of Chinese AI are just as good as those of Western AIs." No wonder, cause the Chinese steal the results.

China’s wholesale inflation tops estimates in August on commodity costs, tech demand as consumer price increases meet forecast

"Much of the pickup reflects a favorable base-effect comparison and higher commodity costs, economists said, rather than a genuine strengthening in household demand [...] The factory-gate inflation was largely concentrated in energy-related sectors, while that for consumer goods continued to fall"

--- Again, real deflation actually continues. There is no proper factory gate inflation. There is simply not enough consumer demand.

"China’s domestic economy remains stuck in a slump, with a negative feedback loop of falling home prices, high savings, weak employment, and slow consumer spending"

--- Yep.

EU's largest political group doesn't want confrontation with China

"The EU has set October as the deadline to secure concessions from the Chinese government, after which point the European Commission is ready to impose "harsher measures", according to EU Trade Chief Maroš Šefčovič. In this context, EPP MEPs sitting on the International Trade Committee are preparing a position paper due to be finalised in mid-October, on the heels of the next European Council summit. The paper focuses on four main areas: rebalancing trade and investment, export control measures, intellectual property rights, and reform of the World Trade Organisation. An early version argues for a balanced approach: promoting dialogue with China while more forcefully defending Europe's economic interests, notably by using trade defence instruments more decisively and developing new tools to counter unfair competition. At the same time, EPP lawmakers insist trade measures should be proportionate and temporary, designed to strengthen the EU's bargaining power rather than lock in permanent protectionism."

--- They are still afraid of China, these idiots. But the thing is, as soon as you start pushing back, China will also push again. And then you should not retreat. That would be stupid, because the only thing the CCP understands is power, strength.

EU unveils ‘buy European’ public procurement rules to counter China

"The new rules will oblige public buyers in the EU to put more emphasis on quality and metrics such as environmental sustainability and local supply chains, rather than opting for the cheapest bid. Under the proposals public agencies will be obliged to give a minimum 30% weighting to quality when they draw up conditions to award a contract, or explain why they chose to disregard this indicator. Labour-intensive contracts will have a minimum 50% weighting for quality, so purchasers must ensure decent jobs when awarding the contract, not simply the cheapest price."

--- Good idea, because particularly in Germany, but not only there, companies and public institutions all too often just choose cheap.

Shifting US alliances push rivals China and India together

"Chinese leader Xi Jinping is visiting New Delhi this weekend, delivering the strongest sign yet of a thaw between the two nuclear-armed neighbors. Behind the tentative steps by Xi and Indian Prime Minister Narendra Modi to ease tensions, analysts say, is the uncertainty in each country’s own relationship with the U.S. [...] One of the clearest signs of that re-evaluation came last year, when Modi traveled to China for the first time in seven years for a regional forum. There, Modi, Xi and Russian leader Vladimir Putin pledged mutual cooperation."

--- The same bullshit as every year. There will be no rapprochement between India & China, at least not on a larger scale, because there are just too many issues between them.

"the repairs are largely cosmetic and won’t alter the fundamental nature of their relationship, especially considering troublesome issues like their disputed border, China’s deep relationship with India’s enemy Pakistan and security competition in the Indian Ocean."

--- Exactly.

China angered by Taiwan VP's visit to EU democracy forum

China-Philippines tensions simmer after Beijing rebukes Manila over Seoul clash

Hong Kong’s Tiananmen vigil organizers handed lengthy prison terms

--- China Uncensored: "Canada's Secret Talks With China's Military"

--- China Uncensored: "China Publicly Threatened the Philippines. It Backfired."

--- news.com.au: Awkward moment Chinese diplomat’s secret note is read out loud in front of live audience

--- Hoover Institution: Why China's Economic Model Is Broken?

Hunderte Visaanträge aus China abgelehnt

"Die Zahl der abgelehnten Visa für Studien- und Forschungsaufenthalte von Chinesen in Deutschland erreicht laut Sicherheitskreisen jährlich einen dreistelligen Wert: Im Jahr 2025 sollen es etwa 200 abgelehnte Anträge von Forschern und Wissenschaftlern bei 2100 erfolgreichen Anträgen gewesen sein. Dazu kommen die abgelehnten Anträge von Studenten: Auch diese sollen sich 2025 in einer Größenordnung von insgesamt rund 200 abgelehnten Anträgen bei insgesamt rund 9000 Anträgen befunden haben. Insgesamt wurden im vergangenen Jahr also etwa 400 Visa verweigert."

--- Bei der Tagesschau klingt es so, als wenn das sehr viele Verweigerungen sind. Sind es nicht. Nicht mal 10%. Da hätte ich mit durchaus deutlich mehr gerechnet.

"Verfassungsschutz-Präsident Sinan Selen erklärte Ende 2025 in einer Anhörung im Bundestag, dass er "Optimierungspotenzial" beim Sicherheitsbewusstsein sehe. Bislang gebe es "unzureichende Abwehrmaßnahmen" beim Screening von Personen, die bei "sensiblen Forschungsvorhaben" mitmachten. Das Problem werde größer, warnte er: "Wir rechnen mit einer deutlichen Zunahme in diesem Bereich." Eine neue "Plattform zur Forschungssicherheit" soll helfen, besser mit möglicher Wissenschaftsspionage umzugehen - ähnliche Einrichtungen gibt es bereits in den Niederlanden oder Großbritannien. Start soll diesen Herbst sein. Man werde mit "einer ersten Sitzung des Lenkungsgremiums” und einem "Soft Opening" - also einem langsamen Hochfahren - der geplanten Servicestelle starten, hieß es aus dem Forschungsministerium. Viele Entscheidungen würden damit weiter von den Universitäten getroffen werden, heißt es aus Wissenschaftskreisen. Aktuell befinde man sich laut Ministerium in der "Feinkonzeptionierung" - genauere Angaben wollte man nicht machen."

--- Wird auch langsam Zeit.

"Aus der chinesischen Botschaft in Berlin hieß es, Sicherheitsfragen sollten "im angemessenen Rahmen" behandelt werden - und nicht als "politisches Instrument" missbraucht werden. Wissenschaft und Technologie gehörten zum "gemeinsamen Reichtum der Menschheit". Die Vorwürfe deutscher Behörden gegen China mit Blick auf Spionageaktivitäten jedenfalls "entbehren jeder Grundlage""

US-Firma Anthropic wirft Konkurrenz aus China Betrug vor

"Die KI-Firma Anthropic wirft chinesischen Rivalen vor, Anfragen von Nutzern heimlich zu ihr umgeleitet zu haben, statt sie mit der eigenen Software zu beantworten. Dadurch seien zum Teil auch Daten staatlicher Stellen aus China und Russland in Systemen der US-Firma gelandet, schrieb Anthropic in einem ausführlichen Bericht zu Versuchen unerlaubter Nutzung. So habe jemand Aufnahmen von Überwachungskameras, unter anderem aus der Umgebung von Standorten der chinesischen Armee, hochgeladen, mit dem Ziel, das Verhalten einer bestimmten Person in den Videos zu analysieren. Der Nutzer ging dabei davon aus, dass die Anfrage von der chinesischen KI-Software Kimi der Firma Moonshot ausgeführt wird. Kimi ließ sie jedoch von Anthropics Claude erledigen, wie es in dem Bericht hieß."

--- Warum sind die Ergebnisse chinesischer KI genauso gut wie die US-amerikanischer KI? Weil die Ergebnisse gestohlen wurden. Überraschung!

"Die Vorwürfe könnten einen Schatten auf das Ansehen der chinesischen KI-Branche werfen. [...] Anleger zweifelten zeitweise an den Geschäftsaussichten amerikanischer KI-Labore wie Anthropic und des ChatGPT-Entwicklers OpenAI, weil der Eindruck entstand, dass leistungsstarke Künstliche Intelligenz mit deutlich weniger Aufwand trainiert werden könne."

--- Die meisten Anleger scheinen Idioten zu sein, die chinesischer Propaganda einfach so Glauben schenken.

Xi gibt "wichtige Anweisungen": 25 Menschen sterben bei Brand auf Frachter in China

"Chinas Staats- und Parteichef Xi Jinping gab laut der amtlichen Nachrichtenagentur Xinhua "wichtige Anweisungen", dass die Suche und Rettung von Vermissten vorangetrieben werden müsse. In der Regel meldet sich Xi nur bei großen Unglücken zu Wort und fordert, wie auch im Fall in Qingdao, wichtige Lehren aus den Vorfällen zu ziehen."

--- Seltsam. Normalerweise würde Kaiser Xi sich bei so einer Situation nicht einmischen. Es sei denn, es gibt wesentlich mehr Opfer. Oder es handelt sich nicht um ein son unwichtiges Frachtschiff, wie es scheint.