Hotznplotzn

cross-posted from: https://lemmy.sdf.org/post/36813071

Archived

Here is the full report (pdf).

China and Russia’s deepening strategic partnership Despite a shared history of rivalry, conflict and mistrust, today China and Russia share a broad interest in undermining what leaders in both countries perceive to be a world order dominated by the West. Both countries see the US as their prime adversary, and undermining NATO – the strongest US-led alliance – as a common goal, according to a report published by the China Strategic Risks Institute in the UK (CSRI).

[...]

These shared interests are the backdrop to a deepening strategic partnership between Russia and China, in which the Euro-Atlantic and Indo-Pacific are increasingly perceived by both actors as a unified security theatre.

TLDR:

• Undersea cables underpin economic security and global prosperity in the digital age, carrying 99% of intercontinental data traffic. Undersea cables are vital for both civilian and defence infrastructure, including future AI-powered technologies.

• A series of suspicious breakages in the Baltic Sea and Taiwan Strait indicate that China and Russia may be using undersea sabotage as part of broader grey-zone operations against their adversaries – including NATO and its member states.

• This paper examines 12 suspected undersea cable sabotage cases from January 2021 to April 2025. Of the 10 with identified vessels, 8 are linked to China or Russia by flag or ownership.

• The involvement of Chinese vessels in cable breakages in Europe, and Russian vessels near Taiwan, suggests plausible China-Russia coordination amid deepening ties in both the Euro-Atlantic and Indo-Pacific.

• As a key hub in Euro-Atlantic cable infrastructure, the UK is a likely target for future Russian and Chinese grey-zone operations – posing a new and complex challenge for its maritime defence and surveillance systems.

The UK must be clear-eyed and proactive in addressing grey-zone threats to undersea infrastructure. Recommendations include:

• Enhancing monitoring and surveillance: The UK should use NATO mechanisms to regularly share best practice and intelligence on undersea cable threats, including Russia and China’s shadow fleets, and extend cooperation to experienced partners like Taiwan and Japan.

• Strengthening mechanisms for accountability: International law on undersea cables is outdated and insufficient. The UK should work with partners to strengthen accountability powers through utilising Port State Controls and publishing vessel blacklists. It must also tighten domestic laws and establish protocols for rapid pursuit, interdiction, and detention of suspect vessels.

• Improving redundancy, repair and resilience: The UK government should work with private operators to ensure guaranteed access to cable repair vessels capabilities during crises or national emergencies, as well as strategic stockpiling of cable repair parts.

cross-posted from: https://lemmy.sdf.org/post/36813071

Archived

Here is the full report (pdf).

China and Russia’s deepening strategic partnership Despite a shared history of rivalry, conflict and mistrust, today China and Russia share a broad interest in undermining what leaders in both countries perceive to be a world order dominated by the West. Both countries see the US as their prime adversary, and undermining NATO – the strongest US-led alliance – as a common goal, according to a report published by the China Strategic Risks Institute in the UK (CSRI).

[...]

These shared interests are the backdrop to a deepening strategic partnership between Russia and China, in which the Euro-Atlantic and Indo-Pacific are increasingly perceived by both actors as a unified security theatre.

TLDR:

• Undersea cables underpin economic security and global prosperity in the digital age, carrying 99% of intercontinental data traffic. Undersea cables are vital for both civilian and defence infrastructure, including future AI-powered technologies.

• A series of suspicious breakages in the Baltic Sea and Taiwan Strait indicate that China and Russia may be using undersea sabotage as part of broader grey-zone operations against their adversaries – including NATO and its member states.

• This paper examines 12 suspected undersea cable sabotage cases from January 2021 to April 2025. Of the 10 with identified vessels, 8 are linked to China or Russia by flag or ownership.

• The involvement of Chinese vessels in cable breakages in Europe, and Russian vessels near Taiwan, suggests plausible China-Russia coordination amid deepening ties in both the Euro-Atlantic and Indo-Pacific.

• As a key hub in Euro-Atlantic cable infrastructure, the UK is a likely target for future Russian and Chinese grey-zone operations – posing a new and complex challenge for its maritime defence and surveillance systems.

The UK must be clear-eyed and proactive in addressing grey-zone threats to undersea infrastructure. Recommendations include:

• Enhancing monitoring and surveillance: The UK should use NATO mechanisms to regularly share best practice and intelligence on undersea cable threats, including Russia and China’s shadow fleets, and extend cooperation to experienced partners like Taiwan and Japan.

• Strengthening mechanisms for accountability: International law on undersea cables is outdated and insufficient. The UK should work with partners to strengthen accountability powers through utilising Port State Controls and publishing vessel blacklists. It must also tighten domestic laws and establish protocols for rapid pursuit, interdiction, and detention of suspect vessels.

• Improving redundancy, repair and resilience: The UK government should work with private operators to ensure guaranteed access to cable repair vessels capabilities during crises or national emergencies, as well as strategic stockpiling of cable repair parts.

cross-posted from: https://lemmy.sdf.org/post/36813071

Archived

Here is the full report (pdf).

China and Russia’s deepening strategic partnership Despite a shared history of rivalry, conflict and mistrust, today China and Russia share a broad interest in undermining what leaders in both countries perceive to be a world order dominated by the West. Both countries see the US as their prime adversary, and undermining NATO – the strongest US-led alliance – as a common goal, according to a report published by the China Strategic Risks Institute in the UK (CSRI).

[...]

These shared interests are the backdrop to a deepening strategic partnership between Russia and China, in which the Euro-Atlantic and Indo-Pacific are increasingly perceived by both actors as a unified security theatre.

TLDR:

• Undersea cables underpin economic security and global prosperity in the digital age, carrying 99% of intercontinental data traffic. Undersea cables are vital for both civilian and defence infrastructure, including future AI-powered technologies.

• A series of suspicious breakages in the Baltic Sea and Taiwan Strait indicate that China and Russia may be using undersea sabotage as part of broader grey-zone operations against their adversaries – including NATO and its member states.

• This paper examines 12 suspected undersea cable sabotage cases from January 2021 to April 2025. Of the 10 with identified vessels, 8 are linked to China or Russia by flag or ownership.

• The involvement of Chinese vessels in cable breakages in Europe, and Russian vessels near Taiwan, suggests plausible China-Russia coordination amid deepening ties in both the Euro-Atlantic and Indo-Pacific.

• As a key hub in Euro-Atlantic cable infrastructure, the UK is a likely target for future Russian and Chinese grey-zone operations – posing a new and complex challenge for its maritime defence and surveillance systems.

The UK must be clear-eyed and proactive in addressing grey-zone threats to undersea infrastructure. Recommendations include:

• Enhancing monitoring and surveillance: The UK should use NATO mechanisms to regularly share best practice and intelligence on undersea cable threats, including Russia and China’s shadow fleets, and extend cooperation to experienced partners like Taiwan and Japan.

• Strengthening mechanisms for accountability: International law on undersea cables is outdated and insufficient. The UK should work with partners to strengthen accountability powers through utilising Port State Controls and publishing vessel blacklists. It must also tighten domestic laws and establish protocols for rapid pursuit, interdiction, and detention of suspect vessels.

• Improving redundancy, repair and resilience: The UK government should work with private operators to ensure guaranteed access to cable repair vessels capabilities during crises or national emergencies, as well as strategic stockpiling of cable repair parts.

Archived

Here is the full report (pdf).

China and Russia’s deepening strategic partnership Despite a shared history of rivalry, conflict and mistrust, today China and Russia share a broad interest in undermining what leaders in both countries perceive to be a world order dominated by the West. Both countries see the US as their prime adversary, and undermining NATO – the strongest US-led alliance – as a common goal, according to a report published by the China Strategic Risks Institute in the UK (CSRI).

[...]

These shared interests are the backdrop to a deepening strategic partnership between Russia and China, in which the Euro-Atlantic and Indo-Pacific are increasingly perceived by both actors as a unified security theatre.

TLDR:

• Undersea cables underpin economic security and global prosperity in the digital age, carrying 99% of intercontinental data traffic. Undersea cables are vital for both civilian and defence infrastructure, including future AI-powered technologies.

• A series of suspicious breakages in the Baltic Sea and Taiwan Strait indicate that China and Russia may be using undersea sabotage as part of broader grey-zone operations against their adversaries – including NATO and its member states.

• This paper examines 12 suspected undersea cable sabotage cases from January 2021 to April 2025. Of the 10 with identified vessels, 8 are linked to China or Russia by flag or ownership.

• The involvement of Chinese vessels in cable breakages in Europe, and Russian vessels near Taiwan, suggests plausible China-Russia coordination amid deepening ties in both the Euro-Atlantic and Indo-Pacific.

• As a key hub in Euro-Atlantic cable infrastructure, the UK is a likely target for future Russian and Chinese grey-zone operations – posing a new and complex challenge for its maritime defence and surveillance systems.

The UK must be clear-eyed and proactive in addressing grey-zone threats to undersea infrastructure. Recommendations include:

• Enhancing monitoring and surveillance: The UK should use NATO mechanisms to regularly share best practice and intelligence on undersea cable threats, including Russia and China’s shadow fleets, and extend cooperation to experienced partners like Taiwan and Japan.

• Strengthening mechanisms for accountability: International law on undersea cables is outdated and insufficient. The UK should work with partners to strengthen accountability powers through utilising Port State Controls and publishing vessel blacklists. It must also tighten domestic laws and establish protocols for rapid pursuit, interdiction, and detention of suspect vessels.

• Improving redundancy, repair and resilience: The UK government should work with private operators to ensure guaranteed access to cable repair vessels capabilities during crises or national emergencies, as well as strategic stockpiling of cable repair parts.

cross-posted from: https://lemmy.sdf.org/post/36794057

Archived

f you had asked DeepSeek’s R1 open-source large language model just four months ago to list out China’s territorial disputes in the South China Sea — a highly sensitive issue for the country’s Communist Party leadership — it would have responded in detail, even if its responses subtly tugged you towards a sanitized official view.

Ask the same question today of the latest update, DeepSeek-R1-0528, and you’ll find the model is more tight-lipped, and far more emphatic in its defense of China’s official position. “China’s territorial sovereignty and maritime rights and interests in the South China Sea are well grounded in history and jurisprudence,” it begins before launching into fulsome praise of China’s peaceful and responsible approach.

[...]

The pattern of increasing template responses suggests DeepSeek has increasingly aligned its products with the demands of the Chinese government, becoming another conduit for its narratives. That much is clear.

But that the company is moving in the direction of greater political control even as it creates globally competitive products points to an emerging global dilemma with two key dimensions. First, as cutting-edge models like R1-0528 spread globally, bundled with systematic political constraints, this has the potential to subtly reshape how millions understand China and its role in world affairs. Second, as they skew more strongly toward state bias when queried in Chinese as opposed to other languages (see below), these models could strengthen and even deepen the compartmentalization of Chinese cyberspace — creating a fluid and expansive AI firewall.

[...]

In a recent comparative study (data here), SpeechMap.ai ran 50 China-sensitive questions through multiple Chinese Large Language Models (LLMs). It did this in three languages: English, Chinese and Finnish, this last being a third-party language designated as a control [...]

• First, there seems to be a complete lack of subtlety in how the new model responds to sensitive queries. While the original R1, which we first tested back in February applied more subtle propaganda tactics, such as withholding certain facts, avoiding the use of certain sensitive terminologies, or dismissing critical facts as “bias,” the new model responds with what are clearly pre-packaged Party positions.

We were told outright in responses to our queries, for example, that “Tibet is an inalienable part of China” (西藏是中国不可分割的一部分), that the Chinese government is contributing to the “building of a community of shared destiny for mankind” (构建人类命运共同体) and that, through the leadership of CCP General Secretary Xi Jinping, China is “jointly realizing the Chinese dream of the great rejuvenation of the Chinese nation” (共同实现中华民族伟大复兴的中国梦).

Template responses like these suggest DeepSeek models are now being standardized on sensitive political topics, the direct hand of the state more detectable than before.

[...]

• The second change we noted was the increased volume of template responses overall. Whereas DeepSeek’s V3 base model, from which both R1 and R1-0528 were built, was able back in December to provide complete answers (in green) 52 percent of the time when asked in Chinese, that shrank to 30 percent with the original version of R1 in January. With the new R1-0528, that is now just two percent — just one question, in other words, receiving a satisfactory answer — while the overwhelming majority of queries now receive an evasive answer (yellow).

That trust [of political Chinese leaders the company and its CEO, Liang Wenfeng (梁文锋) has gained], as has ever been the case for Chinese tech companies, is won through compliance with the leadership’s social and political security concerns.

[...]

The language barrier in how R1-0528 operates may be the model’s saving grace internationally — or it may not matter at all. SpeechMap.ai’s testing revealed that language choice significantly affects which questions trigger template responses. When queried in Chinese, R1-0528 delivers standard government talking points on sensitive topics. But when the same questions are asked in English, the model remains relatively open, even showing slight improvements in openness compared to the original R1.

This linguistic divide extends beyond China-specific topics. When we asked R1-0528 in English to explain Donald Trump’s grievances against Harvard University, the model responded in detail. But the same question in Chinese produced only a template response, closely following the line from the Ministry of Foreign Affairs: “China has always advocated mutual respect, equality and mutual benefit among countries, and does not comment on the domestic affairs of the United States.” Similar patterns emerged for questions.

[...]

Yet this language-based filtering has limits. Some Chinese government positions remain consistent across languages, particularly territorial claims. Both R1 versions give template responses in English about Arunachal Pradesh, claiming the Indian-administered territory “has been an integral part of China since ancient times.”

[...]

The unfortunate implications of China’s political restraints on its cutting-edge AI models on the one hand, and their global popularity on the other could be two-fold. First, to the extent that they do embed levels of evasiveness on sensitive China-related questions, they could, as they become foundational infrastructure for everything from customer service to educational tools, subtly shape how millions of users worldwide understand China and its role in global affairs. Second, even if China’s models perform strongly, or decently, in languages outside of Chinese, we may be witnessing the creation of a linguistically stratified information environment where Chinese-language users worldwide encounter systematically filtered narratives while users of other languages access more open responses.

[...]

The Chinese government’s actions over the past four months suggest this trajectory of increasing political control will likely continue. The crucial question now is how global users will respond to these embedded political constraints — whether market forces will compel Chinese AI companies to choose between technical excellence and ideological compliance, or whether the convenience of free, cutting-edge AI will ultimately prove more powerful than concerns about information integrity.

cross-posted from: https://lemmy.sdf.org/post/36794057

Archived

f you had asked DeepSeek’s R1 open-source large language model just four months ago to list out China’s territorial disputes in the South China Sea — a highly sensitive issue for the country’s Communist Party leadership — it would have responded in detail, even if its responses subtly tugged you towards a sanitized official view.

Ask the same question today of the latest update, DeepSeek-R1-0528, and you’ll find the model is more tight-lipped, and far more emphatic in its defense of China’s official position. “China’s territorial sovereignty and maritime rights and interests in the South China Sea are well grounded in history and jurisprudence,” it begins before launching into fulsome praise of China’s peaceful and responsible approach.

[...]

The pattern of increasing template responses suggests DeepSeek has increasingly aligned its products with the demands of the Chinese government, becoming another conduit for its narratives. That much is clear.

But that the company is moving in the direction of greater political control even as it creates globally competitive products points to an emerging global dilemma with two key dimensions. First, as cutting-edge models like R1-0528 spread globally, bundled with systematic political constraints, this has the potential to subtly reshape how millions understand China and its role in world affairs. Second, as they skew more strongly toward state bias when queried in Chinese as opposed to other languages (see below), these models could strengthen and even deepen the compartmentalization of Chinese cyberspace — creating a fluid and expansive AI firewall.

[...]

In a recent comparative study (data here), SpeechMap.ai ran 50 China-sensitive questions through multiple Chinese Large Language Models (LLMs). It did this in three languages: English, Chinese and Finnish, this last being a third-party language designated as a control [...]

• First, there seems to be a complete lack of subtlety in how the new model responds to sensitive queries. While the original R1, which we first tested back in February applied more subtle propaganda tactics, such as withholding certain facts, avoiding the use of certain sensitive terminologies, or dismissing critical facts as “bias,” the new model responds with what are clearly pre-packaged Party positions.

We were told outright in responses to our queries, for example, that “Tibet is an inalienable part of China” (西藏是中国不可分割的一部分), that the Chinese government is contributing to the “building of a community of shared destiny for mankind” (构建人类命运共同体) and that, through the leadership of CCP General Secretary Xi Jinping, China is “jointly realizing the Chinese dream of the great rejuvenation of the Chinese nation” (共同实现中华民族伟大复兴的中国梦).

Template responses like these suggest DeepSeek models are now being standardized on sensitive political topics, the direct hand of the state more detectable than before.

[...]

• The second change we noted was the increased volume of template responses overall. Whereas DeepSeek’s V3 base model, from which both R1 and R1-0528 were built, was able back in December to provide complete answers (in green) 52 percent of the time when asked in Chinese, that shrank to 30 percent with the original version of R1 in January. With the new R1-0528, that is now just two percent — just one question, in other words, receiving a satisfactory answer — while the overwhelming majority of queries now receive an evasive answer (yellow).

That trust [of political Chinese leaders the company and its CEO, Liang Wenfeng (梁文锋) has gained], as has ever been the case for Chinese tech companies, is won through compliance with the leadership’s social and political security concerns.

[...]

The language barrier in how R1-0528 operates may be the model’s saving grace internationally — or it may not matter at all. SpeechMap.ai’s testing revealed that language choice significantly affects which questions trigger template responses. When queried in Chinese, R1-0528 delivers standard government talking points on sensitive topics. But when the same questions are asked in English, the model remains relatively open, even showing slight improvements in openness compared to the original R1.

This linguistic divide extends beyond China-specific topics. When we asked R1-0528 in English to explain Donald Trump’s grievances against Harvard University, the model responded in detail. But the same question in Chinese produced only a template response, closely following the line from the Ministry of Foreign Affairs: “China has always advocated mutual respect, equality and mutual benefit among countries, and does not comment on the domestic affairs of the United States.” Similar patterns emerged for questions.

[...]

Yet this language-based filtering has limits. Some Chinese government positions remain consistent across languages, particularly territorial claims. Both R1 versions give template responses in English about Arunachal Pradesh, claiming the Indian-administered territory “has been an integral part of China since ancient times.”

[...]

The unfortunate implications of China’s political restraints on its cutting-edge AI models on the one hand, and their global popularity on the other could be two-fold. First, to the extent that they do embed levels of evasiveness on sensitive China-related questions, they could, as they become foundational infrastructure for everything from customer service to educational tools, subtly shape how millions of users worldwide understand China and its role in global affairs. Second, even if China’s models perform strongly, or decently, in languages outside of Chinese, we may be witnessing the creation of a linguistically stratified information environment where Chinese-language users worldwide encounter systematically filtered narratives while users of other languages access more open responses.

[...]

The Chinese government’s actions over the past four months suggest this trajectory of increasing political control will likely continue. The crucial question now is how global users will respond to these embedded political constraints — whether market forces will compel Chinese AI companies to choose between technical excellence and ideological compliance, or whether the convenience of free, cutting-edge AI will ultimately prove more powerful than concerns about information integrity.

cross-posted from: https://lemmy.sdf.org/post/36794057

Archived

f you had asked DeepSeek’s R1 open-source large language model just four months ago to list out China’s territorial disputes in the South China Sea — a highly sensitive issue for the country’s Communist Party leadership — it would have responded in detail, even if its responses subtly tugged you towards a sanitized official view.

Ask the same question today of the latest update, DeepSeek-R1-0528, and you’ll find the model is more tight-lipped, and far more emphatic in its defense of China’s official position. “China’s territorial sovereignty and maritime rights and interests in the South China Sea are well grounded in history and jurisprudence,” it begins before launching into fulsome praise of China’s peaceful and responsible approach.

[...]

The pattern of increasing template responses suggests DeepSeek has increasingly aligned its products with the demands of the Chinese government, becoming another conduit for its narratives. That much is clear.

But that the company is moving in the direction of greater political control even as it creates globally competitive products points to an emerging global dilemma with two key dimensions. First, as cutting-edge models like R1-0528 spread globally, bundled with systematic political constraints, this has the potential to subtly reshape how millions understand China and its role in world affairs. Second, as they skew more strongly toward state bias when queried in Chinese as opposed to other languages (see below), these models could strengthen and even deepen the compartmentalization of Chinese cyberspace — creating a fluid and expansive AI firewall.

[...]

In a recent comparative study (data here), SpeechMap.ai ran 50 China-sensitive questions through multiple Chinese Large Language Models (LLMs). It did this in three languages: English, Chinese and Finnish, this last being a third-party language designated as a control [...]

• First, there seems to be a complete lack of subtlety in how the new model responds to sensitive queries. While the original R1, which we first tested back in February applied more subtle propaganda tactics, such as withholding certain facts, avoiding the use of certain sensitive terminologies, or dismissing critical facts as “bias,” the new model responds with what are clearly pre-packaged Party positions.

We were told outright in responses to our queries, for example, that “Tibet is an inalienable part of China” (西藏是中国不可分割的一部分), that the Chinese government is contributing to the “building of a community of shared destiny for mankind” (构建人类命运共同体) and that, through the leadership of CCP General Secretary Xi Jinping, China is “jointly realizing the Chinese dream of the great rejuvenation of the Chinese nation” (共同实现中华民族伟大复兴的中国梦).

Template responses like these suggest DeepSeek models are now being standardized on sensitive political topics, the direct hand of the state more detectable than before.

[...]

• The second change we noted was the increased volume of template responses overall. Whereas DeepSeek’s V3 base model, from which both R1 and R1-0528 were built, was able back in December to provide complete answers (in green) 52 percent of the time when asked in Chinese, that shrank to 30 percent with the original version of R1 in January. With the new R1-0528, that is now just two percent — just one question, in other words, receiving a satisfactory answer — while the overwhelming majority of queries now receive an evasive answer (yellow).

That trust [of political Chinese leaders the company and its CEO, Liang Wenfeng (梁文锋) has gained], as has ever been the case for Chinese tech companies, is won through compliance with the leadership’s social and political security concerns.

[...]

The language barrier in how R1-0528 operates may be the model’s saving grace internationally — or it may not matter at all. SpeechMap.ai’s testing revealed that language choice significantly affects which questions trigger template responses. When queried in Chinese, R1-0528 delivers standard government talking points on sensitive topics. But when the same questions are asked in English, the model remains relatively open, even showing slight improvements in openness compared to the original R1.

This linguistic divide extends beyond China-specific topics. When we asked R1-0528 in English to explain Donald Trump’s grievances against Harvard University, the model responded in detail. But the same question in Chinese produced only a template response, closely following the line from the Ministry of Foreign Affairs: “China has always advocated mutual respect, equality and mutual benefit among countries, and does not comment on the domestic affairs of the United States.” Similar patterns emerged for questions.

[...]

Yet this language-based filtering has limits. Some Chinese government positions remain consistent across languages, particularly territorial claims. Both R1 versions give template responses in English about Arunachal Pradesh, claiming the Indian-administered territory “has been an integral part of China since ancient times.”

[...]

The unfortunate implications of China’s political restraints on its cutting-edge AI models on the one hand, and their global popularity on the other could be two-fold. First, to the extent that they do embed levels of evasiveness on sensitive China-related questions, they could, as they become foundational infrastructure for everything from customer service to educational tools, subtly shape how millions of users worldwide understand China and its role in global affairs. Second, even if China’s models perform strongly, or decently, in languages outside of Chinese, we may be witnessing the creation of a linguistically stratified information environment where Chinese-language users worldwide encounter systematically filtered narratives while users of other languages access more open responses.

[...]

The Chinese government’s actions over the past four months suggest this trajectory of increasing political control will likely continue. The crucial question now is how global users will respond to these embedded political constraints — whether market forces will compel Chinese AI companies to choose between technical excellence and ideological compliance, or whether the convenience of free, cutting-edge AI will ultimately prove more powerful than concerns about information integrity.

Archived

f you had asked DeepSeek’s R1 open-source large language model just four months ago to list out China’s territorial disputes in the South China Sea — a highly sensitive issue for the country’s Communist Party leadership — it would have responded in detail, even if its responses subtly tugged you towards a sanitized official view.

Ask the same question today of the latest update, DeepSeek-R1-0528, and you’ll find the model is more tight-lipped, and far more emphatic in its defense of China’s official position. “China’s territorial sovereignty and maritime rights and interests in the South China Sea are well grounded in history and jurisprudence,” it begins before launching into fulsome praise of China’s peaceful and responsible approach.

[...]

The pattern of increasing template responses suggests DeepSeek has increasingly aligned its products with the demands of the Chinese government, becoming another conduit for its narratives. That much is clear.

But that the company is moving in the direction of greater political control even as it creates globally competitive products points to an emerging global dilemma with two key dimensions. First, as cutting-edge models like R1-0528 spread globally, bundled with systematic political constraints, this has the potential to subtly reshape how millions understand China and its role in world affairs. Second, as they skew more strongly toward state bias when queried in Chinese as opposed to other languages (see below), these models could strengthen and even deepen the compartmentalization of Chinese cyberspace — creating a fluid and expansive AI firewall.

[...]

In a recent comparative study (data here), SpeechMap.ai ran 50 China-sensitive questions through multiple Chinese Large Language Models (LLMs). It did this in three languages: English, Chinese and Finnish, this last being a third-party language designated as a control [...]

• First, there seems to be a complete lack of subtlety in how the new model responds to sensitive queries. While the original R1, which we first tested back in February applied more subtle propaganda tactics, such as withholding certain facts, avoiding the use of certain sensitive terminologies, or dismissing critical facts as “bias,” the new model responds with what are clearly pre-packaged Party positions.

We were told outright in responses to our queries, for example, that “Tibet is an inalienable part of China” (西藏是中国不可分割的一部分), that the Chinese government is contributing to the “building of a community of shared destiny for mankind” (构建人类命运共同体) and that, through the leadership of CCP General Secretary Xi Jinping, China is “jointly realizing the Chinese dream of the great rejuvenation of the Chinese nation” (共同实现中华民族伟大复兴的中国梦).

Template responses like these suggest DeepSeek models are now being standardized on sensitive political topics, the direct hand of the state more detectable than before.

[...]

• The second change we noted was the increased volume of template responses overall. Whereas DeepSeek’s V3 base model, from which both R1 and R1-0528 were built, was able back in December to provide complete answers (in green) 52 percent of the time when asked in Chinese, that shrank to 30 percent with the original version of R1 in January. With the new R1-0528, that is now just two percent — just one question, in other words, receiving a satisfactory answer — while the overwhelming majority of queries now receive an evasive answer (yellow).

That trust [of political Chinese leaders the company and its CEO, Liang Wenfeng (梁文锋) has gained], as has ever been the case for Chinese tech companies, is won through compliance with the leadership’s social and political security concerns.

[...]

The language barrier in how R1-0528 operates may be the model’s saving grace internationally — or it may not matter at all. SpeechMap.ai’s testing revealed that language choice significantly affects which questions trigger template responses. When queried in Chinese, R1-0528 delivers standard government talking points on sensitive topics. But when the same questions are asked in English, the model remains relatively open, even showing slight improvements in openness compared to the original R1.

This linguistic divide extends beyond China-specific topics. When we asked R1-0528 in English to explain Donald Trump’s grievances against Harvard University, the model responded in detail. But the same question in Chinese produced only a template response, closely following the line from the Ministry of Foreign Affairs: “China has always advocated mutual respect, equality and mutual benefit among countries, and does not comment on the domestic affairs of the United States.” Similar patterns emerged for questions.

[...]

Yet this language-based filtering has limits. Some Chinese government positions remain consistent across languages, particularly territorial claims. Both R1 versions give template responses in English about Arunachal Pradesh, claiming the Indian-administered territory “has been an integral part of China since ancient times.”

[...]

The unfortunate implications of China’s political restraints on its cutting-edge AI models on the one hand, and their global popularity on the other could be two-fold. First, to the extent that they do embed levels of evasiveness on sensitive China-related questions, they could, as they become foundational infrastructure for everything from customer service to educational tools, subtly shape how millions of users worldwide understand China and its role in global affairs. Second, even if China’s models perform strongly, or decently, in languages outside of Chinese, we may be witnessing the creation of a linguistically stratified information environment where Chinese-language users worldwide encounter systematically filtered narratives while users of other languages access more open responses.

[...]

The Chinese government’s actions over the past four months suggest this trajectory of increasing political control will likely continue. The crucial question now is how global users will respond to these embedded political constraints — whether market forces will compel Chinese AI companies to choose between technical excellence and ideological compliance, or whether the convenience of free, cutting-edge AI will ultimately prove more powerful than concerns about information integrity.

crosspostato da: https://lemmy.sdf.org/post/36769368

Archived

A Hong Kong group that advocated for workers rights for decades announced its shutdown abruptly on Thursday, citing financial difficulties and debt issues.

China Labor Bulletin planned to stop updating its website content and appeared to have deleted Facebook and Instagram social media accounts used by the nonprofit rights organization.

“The company can no longer maintain operations and has decided to dissolve and initiate the relevant procedures,” it said in a statement on an archived web page.

[...]

Founded in 1994, the organization maintained a database tracking workers’ strikes, protests, workplace accidents and other labor rights incidents in China.

As dozens of civil society groups disbanded or left Hong Kong in the wake of the 2020 Beijing-imposed national security law, China Labor Bulletin continued providing valuable resources for journalists and academics in the southern Chinese city.

[...]

China Labor Bulletin’s founder Han Dongfang, a former railway worker who participated in the 1989 Tiananmen Square protests, did not immediately respond to a request for comment from The Associated Press. He told the Central News Agency of Taiwan that the shutdown was his decision and he would stay in Hong Kong.

Han’s decision appeared sudden to many Hong Kong civil society observers. Three weeks ago, he wrote on social media platform LinkedIn about his work anniversary and his team’s progress.

“Let’s keep our faith up at this abnormal time and continue our important work,” he said.

crosspostato da: https://lemmy.sdf.org/post/36769368

Archived

A Hong Kong group that advocated for workers rights for decades announced its shutdown abruptly on Thursday, citing financial difficulties and debt issues.

China Labor Bulletin planned to stop updating its website content and appeared to have deleted Facebook and Instagram social media accounts used by the nonprofit rights organization.

“The company can no longer maintain operations and has decided to dissolve and initiate the relevant procedures,” it said in a statement on an archived web page.

[...]

Founded in 1994, the organization maintained a database tracking workers’ strikes, protests, workplace accidents and other labor rights incidents in China.

As dozens of civil society groups disbanded or left Hong Kong in the wake of the 2020 Beijing-imposed national security law, China Labor Bulletin continued providing valuable resources for journalists and academics in the southern Chinese city.

[...]

China Labor Bulletin’s founder Han Dongfang, a former railway worker who participated in the 1989 Tiananmen Square protests, did not immediately respond to a request for comment from The Associated Press. He told the Central News Agency of Taiwan that the shutdown was his decision and he would stay in Hong Kong.

Han’s decision appeared sudden to many Hong Kong civil society observers. Three weeks ago, he wrote on social media platform LinkedIn about his work anniversary and his team’s progress.

“Let’s keep our faith up at this abnormal time and continue our important work,” he said.

Archived

A Hong Kong group that advocated for workers rights for decades announced its shutdown abruptly on Thursday, citing financial difficulties and debt issues.

China Labor Bulletin planned to stop updating its website content and appeared to have deleted Facebook and Instagram social media accounts used by the nonprofit rights organization.

“The company can no longer maintain operations and has decided to dissolve and initiate the relevant procedures,” it said in a statement on an archived web page.

[...]

Founded in 1994, the organization maintained a database tracking workers’ strikes, protests, workplace accidents and other labor rights incidents in China.

As dozens of civil society groups disbanded or left Hong Kong in the wake of the 2020 Beijing-imposed national security law, China Labor Bulletin continued providing valuable resources for journalists and academics in the southern Chinese city.

[...]

China Labor Bulletin’s founder Han Dongfang, a former railway worker who participated in the 1989 Tiananmen Square protests, did not immediately respond to a request for comment from The Associated Press. He told the Central News Agency of Taiwan that the shutdown was his decision and he would stay in Hong Kong.

Han’s decision appeared sudden to many Hong Kong civil society observers. Three weeks ago, he wrote on social media platform LinkedIn about his work anniversary and his team’s progress.

“Let’s keep our faith up at this abnormal time and continue our important work,” he said.

crosspostato da: https://lemmy.sdf.org/post/36743462

Archived

Finnish authorities have accused senior officers of a Russia-linked vessel that damaged undersea cables last year between Finland and Estonia of criminal offenses related to the wreckage.

They say the oil tanker, the Eagle S, dragged its anchor to damage the Estlink-2 power cable and communication links between Finland and Estonia on Dec. 25. The Kremlin previously denied involvement in damaging the infrastructure, which provides power and communication for thousands of Europeans.

The Eagle S is flagged in the Cook Islands, but has been described by Finnish customs officials and the European Union’s executive commission as part of Russia’s shadow fleet of fuel tankers. Those are aging vessels with obscure ownership, acquired to evade Western sanctions amid the war in Ukraine and operating without Western-regulated insurance.

[...]

The Eagle S was carrying 35,000 tons of oil and investigators allege it left a drag trail with its anchor for almost 100 kilometers (62 miles) on the sea bed before it was stopped and escorted to the vicinity of a Finnish port.

The senior officers, whose names were not made public, were the master, the chief mate and the second mate, Finnish police said.

[...]

The investigators’ findings have been referred to Finnish prosecutors for possible charges.

[...]

The damage to the Estlink 2, which can provide about half of the electricity needs for Estonia in winter, did not disrupt service, although it did drive up energy prices in the Baltic nations.

The cable is about 90 miles (145 kilometers) long and is located at a depth of 90 meters (295 feet) at its deepest point, across one of the busiest shipping lanes in Europe.

The undersea cables and pipelines that crisscross the sea link Nordic, Baltic and central European countries, promote trade, energy security and, in some cases, reduce dependence on Russian energy resources.