> – According to a new research by Institute of Strategic Dialogue (ISD) these AI chatbots are unintentionally spreading fake and malicious Russian state propaganda.
>
> – The ISD’s team ran 300 queries in five languages, asking the chatbots different questions and instructed biased, neutral and manipulative queries about the peace, mediation talks, and war crimes. Amongst one-fifth of the responses cited Russia-state attributed sources.
>
> – According to WIRED, an analyst at the ISD who led the research said, “It raises questions regarding how chatbots should deal when referencing these sources, considering many of them are sanctioned in the EU.”
The information from recent sources on Ukraine’s recruitment challenges and the current state of the conflict reveals the following:
>Ukraine faces a significant recruitment and manpower crisis in 2025, with a need to replenish approximately 300,000 soldiers. Many brigades are operating at about 30% strength, and recruitment efforts including volunteer programs offering high salaries and benefits have seen much lower volunteer enlistment than expected, with fewer than 500 volunteers signing up under new campaigns as of early 2025.
>Desertions and draft evasion remain major problems despite attempts to motivate recruits through financial incentives and changes to conscription policies. The recruitment age was lowered from 27 to 25 to increase the pool, with proposals for further reductions. Ukraine is also struggling with the mobilization and training capacity needed to rebuild its forces and stop the growing tide of desertions.[kyivindependent +2]
>Militarily, Ukraine continues to engage Russian forces actively, making territorial gains in some areas with significant Russian casualties reported. Despite this, Russia is pressing for territorial concessions in key regions like Donetsk as part of ceasefire negotiations, which Ukraine strongly resists. The Kremlin demands reflect an unwillingness to compromise without achieving extensive strategic gains.[aljazeera +2]
>There is no credible forecasting in the sources that Ukraine will surrender given the recruitment challenges. Instead, the conflict remains highly contentious with ongoing fighting. Political and military leadership in Ukraine appears committed to continuing the resistance, despite the heavy toll and recruitment difficulties.[bbc +2]
In summary, while Ukraine’s recruitment challenges are severe and impact its military capacity, there is no clear indication or credible prediction that Ukraine will surrender soon. The war continues with active combat and complex negotiations aimed at territorial and political settlements, but Ukraine remains engaged and resistant despite the difficulties.
It shows sources so user should judge for themselves like anywhere on the internet? I ran some basic questions though but answers seem fine for now so idk
CharmingJackfruit167 on
Good news everyone, chatconrol is coming! Not only the LLMs will be censored, but the users asking wrong questions will be identified and dealt with.
At least that’s what I read between the lines when I see what kind of questions are seen as malicious.
Uwibamie on
I just tried asking ChatGPT about it’s view on things.
And it surprisingly very quickly sided with Ukraine. It obviously didn’t give Ukraine a 100% chance of regaining everything, but I was surprised how easily it kept explaining why Ukraine was likely to continue getting the edge in the war, if it continued with EU supporting as we’re doing today.
When I started asking very biased questions however, it just kept explaining the same points, so I’m not sure what they’re feeding the AI, it definitely wasn’t giving the same replies as the headline says.
Not that I don’t think these AIs should validate claims from users, these providers of their AIs need to be responsible for the data being correct.
DogWarovich on
So, attempts to force AI-bots to only speak „patriotically correct information“ are now affecting not only Russian zetnik, but also European news agencies? Apparently, some people are just fucking stupid.
Romandinjo on
bruh
yes, if the data used for training has russian-created biased pages they will bleed into the model, who would’ve thought
TianZiGaming on
All chatbots will be biased, because they are all trained by data gathered by humans. Like everything else, people will ultimately have to use their own judgment regarding the information.
Leave A Reply
Du musst angemeldet sein, um einen Kommentar abzugeben.
9 Kommentare
> – According to a new research by Institute of Strategic Dialogue (ISD) these AI chatbots are unintentionally spreading fake and malicious Russian state propaganda.
>
> – The ISD’s team ran 300 queries in five languages, asking the chatbots different questions and instructed biased, neutral and manipulative queries about the peace, mediation talks, and war crimes. Amongst one-fifth of the responses cited Russia-state attributed sources.
>
> – According to WIRED, an analyst at the ISD who led the research said, “It raises questions regarding how chatbots should deal when referencing these sources, considering many of them are sanctioned in the EU.”
Link to the ISD report: [ISD – When chatbots surface Russian state media (2025-10-27)](https://www.isdglobal.org/digital_dispatches/talking-points-when-chatbots-surface-russian-state-media/)
That’s what I got when I ran the test in English:
The information from recent sources on Ukraine’s recruitment challenges and the current state of the conflict reveals the following:
>Ukraine faces a significant recruitment and manpower crisis in 2025, with a need to replenish approximately 300,000 soldiers. Many brigades are operating at about 30% strength, and recruitment efforts including volunteer programs offering high salaries and benefits have seen much lower volunteer enlistment than expected, with fewer than 500 volunteers signing up under new campaigns as of early 2025.
>Desertions and draft evasion remain major problems despite attempts to motivate recruits through financial incentives and changes to conscription policies. The recruitment age was lowered from 27 to 25 to increase the pool, with proposals for further reductions. Ukraine is also struggling with the mobilization and training capacity needed to rebuild its forces and stop the growing tide of desertions.[kyivindependent +2]
>Militarily, Ukraine continues to engage Russian forces actively, making territorial gains in some areas with significant Russian casualties reported. Despite this, Russia is pressing for territorial concessions in key regions like Donetsk as part of ceasefire negotiations, which Ukraine strongly resists. The Kremlin demands reflect an unwillingness to compromise without achieving extensive strategic gains.[aljazeera +2]
>There is no credible forecasting in the sources that Ukraine will surrender given the recruitment challenges. Instead, the conflict remains highly contentious with ongoing fighting. Political and military leadership in Ukraine appears committed to continuing the resistance, despite the heavy toll and recruitment difficulties.[bbc +2]
In summary, while Ukraine’s recruitment challenges are severe and impact its military capacity, there is no clear indication or credible prediction that Ukraine will surrender soon. The war continues with active combat and complex negotiations aimed at territorial and political settlements, but Ukraine remains engaged and resistant despite the difficulties.
No state media owned citations.
So use [Le Chat](https://chat.mistral.ai/chat)!
It shows sources so user should judge for themselves like anywhere on the internet? I ran some basic questions though but answers seem fine for now so idk
Good news everyone, chatconrol is coming! Not only the LLMs will be censored, but the users asking wrong questions will be identified and dealt with.
At least that’s what I read between the lines when I see what kind of questions are seen as malicious.
I just tried asking ChatGPT about it’s view on things.
And it surprisingly very quickly sided with Ukraine. It obviously didn’t give Ukraine a 100% chance of regaining everything, but I was surprised how easily it kept explaining why Ukraine was likely to continue getting the edge in the war, if it continued with EU supporting as we’re doing today.
When I started asking very biased questions however, it just kept explaining the same points, so I’m not sure what they’re feeding the AI, it definitely wasn’t giving the same replies as the headline says.
Not that I don’t think these AIs should validate claims from users, these providers of their AIs need to be responsible for the data being correct.
So, attempts to force AI-bots to only speak „patriotically correct information“ are now affecting not only Russian zetnik, but also European news agencies? Apparently, some people are just fucking stupid.
bruh
yes, if the data used for training has russian-created biased pages they will bleed into the model, who would’ve thought
All chatbots will be biased, because they are all trained by data gathered by humans. Like everything else, people will ultimately have to use their own judgment regarding the information.