Russia launched massive overnight strike on defense industry sites, depots, transport & logistics hubs in Kiev Region
In Kiev alone, the strikes hit:
- Fire Point — UAVs & missiles
- Ukrspecsystems — UAV
- Antonov Serial Production Plant — aviation
“
That’s how the Russian Defense Ministry commented on the massive overnight strike on targets in Kyiv and the Kyiv region.
In total, Russian air defense systems shot down 726 Ukrainian drones over Russian regions overnight.
@militarysummary & @MTodayNews
#Ukraine
In Kiev alone, the strikes hit:
- Fire Point — UAVs & missiles
- Ukrspecsystems — UAV
- Antonov Serial Production Plant — aviation
“
We’re making a compelling case.”
That’s how the Russian Defense Ministry commented on the massive overnight strike on targets in Kyiv and the Kyiv region.
In total, Russian air defense systems shot down 726 Ukrainian drones over Russian regions overnight.
@militarysummary & @MTodayNews
note : the 4 top pictures are from the previous massive strikes (05/08/2026) https://t.me/militarysummary/30222?single#Ukraine
Forwarded from 0•Bytes•1
ai_myths_en.pdf
4.3 MB
Hello everyone, my Alices and Cheshire cats! 🐈⬛🎀
It's finally done - the article some of you have been waiting for since last year! 🎉
This time I'm debunking the myths that make people believe their conversations with popular AI are at least somewhat "private" 🫥
We'll look at why that's not true, and how neural networks actually collect your data -from "private mode" to chat deletion and premium subscriptions.
But the most important part isn't just to scare you - it's to show you methods that actually work 🛠 I'll walk through a few approaches I've personally tested that genuinely help reduce the risks, including local models that don't send anything outside your device.
I hope you find this topic as interesting to dive into as I did! ♥️
And if you're into AI topics, I might soon put together a separate post or short article on AI-related cybersecurity incidents, since there's a lot happening on that front right now 🫠
#ai #anonymity #local_ai #OPSEC #privacy #neural_networks #data_leaks #tools
It's finally done - the article some of you have been waiting for since last year! 🎉
This time I'm debunking the myths that make people believe their conversations with popular AI are at least somewhat "private" 🫥
We'll look at why that's not true, and how neural networks actually collect your data -from "private mode" to chat deletion and premium subscriptions.
But the most important part isn't just to scare you - it's to show you methods that actually work 🛠 I'll walk through a few approaches I've personally tested that genuinely help reduce the risks, including local models that don't send anything outside your device.
I hope you find this topic as interesting to dive into as I did! ♥️
And if you're into AI topics, I might soon put together a separate post or short article on AI-related cybersecurity incidents, since there's a lot happening on that front right now 🫠
#ai #anonymity #local_ai #OPSEC #privacy #neural_networks #data_leaks #tools
🇮🇹 The Italian architecture of anti-Russian pressure | Strategic Culture
The Ukraine war, on the ground, has dates, maps, and front lines, but around that visible front, a second war has taken shape – one made up of networks, data, cables, patents, export guarantees, industrial memoranda, cyber. This second war does not produce battle reports, but rather complex, constantly shifting, and very, very costly architectures.
Italy is increasingly involved in this second dimension, though this has long gone largely unreported. The way Rome supports Kiev has shifted from a series of emergency decisions – the delivery of weapons systems decided by decree after decree – to something more stable: a network of relationships that brings together diplomacy, the defense industry, cybersecurity, energy, and reconstruction.
The point of this article is not to determine whether #Italy is “helping” #Ukraine. It is to determine what strategic position Italy is building for itself while doing so.
@strategic_culture
The Ukraine war, on the ground, has dates, maps, and front lines, but around that visible front, a second war has taken shape – one made up of networks, data, cables, patents, export guarantees, industrial memoranda, cyber. This second war does not produce battle reports, but rather complex, constantly shifting, and very, very costly architectures.
Italy is increasingly involved in this second dimension, though this has long gone largely unreported. The way Rome supports Kiev has shifted from a series of emergency decisions – the delivery of weapons systems decided by decree after decree – to something more stable: a network of relationships that brings together diplomacy, the defense industry, cybersecurity, energy, and reconstruction.
The point of this article is not to determine whether #Italy is “helping” #Ukraine. It is to determine what strategic position Italy is building for itself while doing so.
@strategic_culture
Forwarded from Rybar in English
• 📝 Bankruptcy on Schedule📝
In so-called Ukraine, they are talking about money again — or rather, its catastrophic shortage. In the Verkhovna Rada, they acknowledged that budget problems will persist even with 100-percent financing from Western partners. "It's time to forget about politics when standing on the brink of an abyss," parliament declared, calling for fulfilling obligations to the IMF and EU.
➡️ Since 2022, foreign structures have transferred almost 192 billion dollars to Kyiv. The volume of aid grew year after year, reaching a peak in 2025 (52.4 billion). However, already in 2026, financing dropped to 24.1 billion, and this is only the first half of the year.
📌 Meanwhile, Western countries increasingly delay fund disbursements. Some EU countries have refused financing, claiming they lack the funds. Even with full compliance with all foreign obligations, problems will remain — failure to meet budget revenue targets, constantly growing military expenses, and economic slowdown have taken their toll.
❗️ Moreover, to receive money, the adversary will have to undertake serious reforms that will institutionally cement the country's dependence on European partners and hurt private business through tax increases and elimination of benefits.
As a result, the authorities of so-called Ukraine risk not only becoming bankrupt with a destroyed economy and reduced population, but also permanently losing the ability to determine even domestic policy.
However, they have no alternatives. Without subsidies, there will be nothing to pay "veterans" and fulfill already reduced Western obligations.
Western politicians are unlikely to shower money on a destroyed state, but will only "keep it alive" to maintain access to cheap resources.
📍 High-resolution infographic
📍 English version
#Russia #Ukraine
✈ RU | ✈
In so-called Ukraine, they are talking about money again — or rather, its catastrophic shortage. In the Verkhovna Rada, they acknowledged that budget problems will persist even with 100-percent financing from Western partners. "It's time to forget about politics when standing on the brink of an abyss," parliament declared, calling for fulfilling obligations to the IMF and EU.
🔻 Who and how much financed:▪️ USA — main grant donor: over 30 billion dollars in the period from 2022 to 2024.▪️ European Union — systemic creditor: since 2023, transferred about 60 billion dollars.▪️ IMF — over 14.8 billion since 2022.▪️ ERA Mechanism (G7) — 47.7 billion in 2024–2026 from frozen Russian assets.
As a result, the authorities of so-called Ukraine risk not only becoming bankrupt with a destroyed economy and reduced population, but also permanently losing the ability to determine even domestic policy.
However, they have no alternatives. Without subsidies, there will be nothing to pay "veterans" and fulfill already reduced Western obligations.
Western politicians are unlikely to shower money on a destroyed state, but will only "keep it alive" to maintain access to cheap resources.
#Russia #Ukraine
Please open Telegram to view this post
VIEW IN TELEGRAM
Graphene OS explained why /e/ os murena and fairphone are bad at
https://grapheneos.social/@GrapheneOS
https://grapheneos.social/@GrapheneOS
GrapheneOS Mastodon
GrapheneOS (@GrapheneOS@grapheneos.social)
18.1K Posts, 0 Following, 38.6K Followers · Open source privacy and security focused mobile OS with Android app compatibility.
And in this thread they explain how goolag are violating the GPL and making it harder and slower to get the source code for android in pixels.
They say this won't happen with their Motorola graphene OS devices starting in 2027
https://grapheneos.social/@GrapheneOS/117078064184215730
#google #Android #GPL #pixel #sources #why
They say this won't happen with their Motorola graphene OS devices starting in 2027
https://grapheneos.social/@GrapheneOS/117078064184215730
#google #Android #GPL #pixel #sources #why
GrapheneOS Mastodon
GrapheneOS (@GrapheneOS@grapheneos.social)
@Revan_9766@mastodon.social The initial devices with GrapheneOS support should be available in 2027. The initial devices will be flagships so they'll be higher end hardware than Pixels at a higher price. Lower end devices will take more time to meet our requirements…
https://nitter.net/GrapheneOS/status/2081837057433915603
/e/ is a highly problematic fork of LineageOS very poor privacy and atrocious security. People would be far better off using standard LineageOS one of the better supported devices. An iPhone is far more private and drastically more secure than Murena products.
/e/ lacks crucial standard privacy and security patches. It has many months and even years of delays to provide those. Nearly all the devices they support lack most Linux kernel, driver and firmware updates. They don't keep up with those for any devices. They don't keep up with Android, Chromium or Linux security updates anywhere either. Many severe patched privacy and security vulnerabilities are left wide open for months and even years with /e/. They heavily mislead their users about what's patched and how long they take to ship it.
/e/ also doesn't provide important standard privacy or security protections. It lags years behind on providing the current generation protections and disables many of the most important ones. An operating system failing to protect against known privacy weaknesses does not have reasonable privacy.
/e/ includes their own privacy invasive services and sends user data to third party including OpenAI without consent. Despite how it's marketed, it includes many Google services with privileged integration into the OS. They have their own privacy invasive services including user tracking with unique identifiers in their update client.
Their own supposed privacy features are incredibly flawed and do not provide what they claim. A small list of blocked domains omitting everything used for useful functionality does not stop tracking by apps or services. It leaves the most privacy invasive behavior of apps and services intact and is trivially bypassed in multiple ways even for what it does cover. Apps and services only need to use the same domains for functionality as tracking to avoid it which is already largely what they do. Apps and services also already widely deploy bypasses for this naive domain-based filtering even for domains dedicated to tracking. /e/ would have you believe that a DNS blocklist is a game changing approach. There are better implementations including RethinkDNS available for use elsewhere.
More information:
community.e.foundation/t/voi…
codeberg.org/divested-mobile…
eylenburg.github.io/android_…
discuss.grapheneos.org/d/241…
/e/ and their for-profit company Murena have repeatedly pushed anti-privacy talking points from authoritarians claiming devices with strong privacy and security are for criminals and pedophiles. Here are 2 examples where Gaël Duval says that himself about privacy/security hardened devices in general:
nitter.net/GrapheneOS/status/2040…
clubic.com/actualite-604786-…
/e/ and Murena many of the same anti-privacy talking points as France's current government and national law enforcement. They've aligned themselves with authoritarians in the country where they're based instead of opposing it. There are dozens of examples where they falsely claim GrapheneOS are mainly useful to criminals and mainly used by criminals. That's more extreme than the claims by France's national law enforcement cracking down on secure devices. They've gone so far to align themselves with anti-privacy authoritarians that they're ahead of them.
Murena is a for-profit company with the founders and owners of it being the same people who run /e/. /e/ includes for-profit paid Murena services and is built for Murena to sell devices with it. Despite being made for profit, /e/ receives millions of euros in government funding from governments which have spent years cracking down on legitimate privacy products:
> The European Union has subsidized us to the tune of several million for this project.
projets-libres.org/en/podcas…
#lel
/e/ is a highly problematic fork of LineageOS very poor privacy and atrocious security. People would be far better off using standard LineageOS one of the better supported devices. An iPhone is far more private and drastically more secure than Murena products.
/e/ lacks crucial standard privacy and security patches. It has many months and even years of delays to provide those. Nearly all the devices they support lack most Linux kernel, driver and firmware updates. They don't keep up with those for any devices. They don't keep up with Android, Chromium or Linux security updates anywhere either. Many severe patched privacy and security vulnerabilities are left wide open for months and even years with /e/. They heavily mislead their users about what's patched and how long they take to ship it.
/e/ also doesn't provide important standard privacy or security protections. It lags years behind on providing the current generation protections and disables many of the most important ones. An operating system failing to protect against known privacy weaknesses does not have reasonable privacy.
/e/ includes their own privacy invasive services and sends user data to third party including OpenAI without consent. Despite how it's marketed, it includes many Google services with privileged integration into the OS. They have their own privacy invasive services including user tracking with unique identifiers in their update client.
Their own supposed privacy features are incredibly flawed and do not provide what they claim. A small list of blocked domains omitting everything used for useful functionality does not stop tracking by apps or services. It leaves the most privacy invasive behavior of apps and services intact and is trivially bypassed in multiple ways even for what it does cover. Apps and services only need to use the same domains for functionality as tracking to avoid it which is already largely what they do. Apps and services also already widely deploy bypasses for this naive domain-based filtering even for domains dedicated to tracking. /e/ would have you believe that a DNS blocklist is a game changing approach. There are better implementations including RethinkDNS available for use elsewhere.
More information:
community.e.foundation/t/voi…
codeberg.org/divested-mobile…
eylenburg.github.io/android_…
discuss.grapheneos.org/d/241…
/e/ and their for-profit company Murena have repeatedly pushed anti-privacy talking points from authoritarians claiming devices with strong privacy and security are for criminals and pedophiles. Here are 2 examples where Gaël Duval says that himself about privacy/security hardened devices in general:
nitter.net/GrapheneOS/status/2040…
clubic.com/actualite-604786-…
/e/ and Murena many of the same anti-privacy talking points as France's current government and national law enforcement. They've aligned themselves with authoritarians in the country where they're based instead of opposing it. There are dozens of examples where they falsely claim GrapheneOS are mainly useful to criminals and mainly used by criminals. That's more extreme than the claims by France's national law enforcement cracking down on secure devices. They've gone so far to align themselves with anti-privacy authoritarians that they're ahead of them.
Murena is a for-profit company with the founders and owners of it being the same people who run /e/. /e/ includes for-profit paid Murena services and is built for Murena to sell devices with it. Despite being made for profit, /e/ receives millions of euros in government funding from governments which have spent years cracking down on legitimate privacy products:
> The European Union has subsidized us to the tune of several million for this project.
projets-libres.org/en/podcas…
#lel
/e/ makes it far easier to remotely take over devices, violate user privacy from apps and bypass encryption to extract all the data from a device. These are exactly the kinds of devices authoritarians opposed to privacy and security want people to be using. These devices are much less private and drastically less secure than iPhones but are duping people into paying them for it with false marketing. They're regressing privacy and security instead of advancing it.
NoGoolag
Video
Gaël Duval is the founder and president of the /e/ foundation along with the CEO of #Murena. #Duval and his organizations have consistently taken a stance against protecting users from exploits. In this video, he once again claims protecting against exploits is for only useful pedophiles and spies.
https://nitter.net/GrapheneOS/status/2040887784253141142
Translation to English:
> There's the attack surface, on that front we're not security specialists here, so I couldn't answer you precisely, but from the discussions I've had, it seems that everything we do reduces attack surface. However, we don't have a "hardened security" approach, we aren't developing a phone for pedo(censored) so they can evade justice. So there aren't difficult things to check if the memory is corrupted, really hardened security stuff that could clearly be useful for executives, in the secret service, or whatever. That's not our goal, our goal is to start from an observation: today our personal data is constantly being plundered and that wouldn't be legal in real life with the mail or the telephone, we want to change that. So we are making you a product that changes that by default for anyone.
Transcription in French:
> Il y a la surface d'attaque, là pour le coup on est pas des spécialistes de la sécurité, donc je ne pourrais pas te répondre avec précision, mais des discussions que j'ai eu, il semblerait que tout ce qu'on fait, ça réduit la surface d'attaque. Donc oui, probablement ça aide. Par contre, on a pas une approche "sécurité durcie", on développe pas un téléphone pour les pédo(bip) pour qu'ils puissent échapper à la justice. Donc il y a pas des trucs pas possibles pour voir si la mémoire est pas corrompue, des trucs de sécu vraiment durcis qui pourraient être utiles clairement pour des dirigeants, dans les services secrets ou que sais-je. C'est pas notre but, notre but c'est de partir d'un constat, aujourd'hui nos données personnelles sont pillées en permanence et ça serait pas légal dans la vraie vie avec le courrier ou le téléphone, on veut changer ça. Donc on vous fait un produit qui change ça par défaut pour n'importe quelle personne.
/e/ and Murena repeatedly claim privacy/security hardened devices are mainly for criminals and pedophiles. They've also been falsely claiming GrapheneOS is mainly used by criminals for years alongside misleading people about what it provides. That's why we began providing our perspective in response. Fairphone gave a weasel worded media statement supporting Murena in this and other involvement.
https://nitter.net/GrapheneOS/status/2040887784253141142
https://www.clubic.com/actualite-604786-murena-e-os-interview.html
https://nitter.net/GrapheneOS/status/2
#lel
https://nitter.net/GrapheneOS/status/2040887784253141142
Translation to English:
> There's the attack surface, on that front we're not security specialists here, so I couldn't answer you precisely, but from the discussions I've had, it seems that everything we do reduces attack surface. However, we don't have a "hardened security" approach, we aren't developing a phone for pedo(censored) so they can evade justice. So there aren't difficult things to check if the memory is corrupted, really hardened security stuff that could clearly be useful for executives, in the secret service, or whatever. That's not our goal, our goal is to start from an observation: today our personal data is constantly being plundered and that wouldn't be legal in real life with the mail or the telephone, we want to change that. So we are making you a product that changes that by default for anyone.
Transcription in French:
> Il y a la surface d'attaque, là pour le coup on est pas des spécialistes de la sécurité, donc je ne pourrais pas te répondre avec précision, mais des discussions que j'ai eu, il semblerait que tout ce qu'on fait, ça réduit la surface d'attaque. Donc oui, probablement ça aide. Par contre, on a pas une approche "sécurité durcie", on développe pas un téléphone pour les pédo(bip) pour qu'ils puissent échapper à la justice. Donc il y a pas des trucs pas possibles pour voir si la mémoire est pas corrompue, des trucs de sécu vraiment durcis qui pourraient être utiles clairement pour des dirigeants, dans les services secrets ou que sais-je. C'est pas notre but, notre but c'est de partir d'un constat, aujourd'hui nos données personnelles sont pillées en permanence et ça serait pas légal dans la vraie vie avec le courrier ou le téléphone, on veut changer ça. Donc on vous fait un produit qui change ça par défaut pour n'importe quelle personne.
/e/ and Murena repeatedly claim privacy/security hardened devices are mainly for criminals and pedophiles. They've also been falsely claiming GrapheneOS is mainly used by criminals for years alongside misleading people about what it provides. That's why we began providing our perspective in response. Fairphone gave a weasel worded media statement supporting Murena in this and other involvement.
https://nitter.net/GrapheneOS/status/2040887784253141142
https://www.clubic.com/actualite-604786-murena-e-os-interview.html
https://nitter.net/GrapheneOS/status/2
#lel
Nitter
GrapheneOS (@GrapheneOS)
Gaël Duval is the founder and president of the /e/ foundation along with the CEO of Murena. Duval and his organizations have consistently taken a stance against protecting users from exploits. In this video, he once again claims protecting against exploits…
#PINE64 is Halting its Linux Hardware Line, and The AI Bubble is to Blame
They are blaming DRAM and eMMC shortages, with nothing coming back before mid-2027.
https://itsfoss.com/news/pine64-kills-off-linux-hardware/
They are blaming DRAM and eMMC shortages, with nothing coming back before mid-2027.
https://itsfoss.com/news/pine64-kills-off-linux-hardware/
It's FOSS
PINE64 is Halting its Linux Hardware Line, and The AI Bubble is to Blame
They are blaming DRAM and eMMC shortages, with nothing coming back before mid-2027.
Forwarded from Unlimited Hangout
NEW from Whitney Webb and Mark Goodwin
Javier Milei’s administration has been showcasing Patagonia to Silicon Valley executives, culminating in OpenAI’s recent ‘Stargate Argentina’ announcement. The focus of this piece, Emiliano Kargieman, is the man actually developing OpenAI’s Argentine project and boasts significant ties to United States’ military and intelligence, NASA, Palantir and the Silicon Valley stalwarts who helped privatize space.
Here, we detail how Kargieman has said the quiet apart loud about how data center construction isn't about "making jobs", it's about making money, and how he has suggested that OpenAI is a front for the accelerated build-out of AI infrastructure (financed by an intentional debt bubble). He has also suggested that Stargate Argentina will eventually be used, not by OpenAI (which Kargieman anticipates will collapse when the AI bubble pops), but by the Argentine State.
In looking at Javier Milei's policies with respect to AI domestically in Argentina, the "digital transformation" of the Argentine State - which Stargate Argentina hopes to further advance - is largely for "predictive" behavioral control of its population. We also touch on some of the likely motivators of Peter Thiel's move to Argentina and where he fits into all of this.
https://unlimitedhangout.com/2026/08/investigative-reports/how-argentinas-elon-musk-is-bringing-technocracy-to-patagonia/
Javier Milei’s administration has been showcasing Patagonia to Silicon Valley executives, culminating in OpenAI’s recent ‘Stargate Argentina’ announcement. The focus of this piece, Emiliano Kargieman, is the man actually developing OpenAI’s Argentine project and boasts significant ties to United States’ military and intelligence, NASA, Palantir and the Silicon Valley stalwarts who helped privatize space.
Here, we detail how Kargieman has said the quiet apart loud about how data center construction isn't about "making jobs", it's about making money, and how he has suggested that OpenAI is a front for the accelerated build-out of AI infrastructure (financed by an intentional debt bubble). He has also suggested that Stargate Argentina will eventually be used, not by OpenAI (which Kargieman anticipates will collapse when the AI bubble pops), but by the Argentine State.
In looking at Javier Milei's policies with respect to AI domestically in Argentina, the "digital transformation" of the Argentine State - which Stargate Argentina hopes to further advance - is largely for "predictive" behavioral control of its population. We also touch on some of the likely motivators of Peter Thiel's move to Argentina and where he fits into all of this.
https://unlimitedhangout.com/2026/08/investigative-reports/how-argentinas-elon-musk-is-bringing-technocracy-to-patagonia/
Unlimited Hangout
How “Argentina’s Elon Musk” Is Bringing Technocracy To Patagonia
Javier Milei’s administration has been showcasing Patagonia to Silicon Valley executives, culminating in OpenAI’s ‘Stargate Argentina’ announcement. The focus of this piece, Emiliano Kargieman, is the man actually developing OpenAI’s Argentine project and…
In case you need to see it to believe it, here is OpenAI's partner in Stargate Argentina saying that AI data centers are not a source of jobs, but a source of dollars. This is despite Argentine politicians, including Milei and others, claiming the opposite.
https://fixupx.com/_whitneywebb/status/2090449946952585289
https://fixupx.com/_whitneywebb/status/2090449946952585289
🧵 Thread • FixupX
Whitney Webb (@_whitneywebb)
↩ (@_whitneywebb)
In case you need to see it to believe it, here is OpenAI's partner in Stargate Argentina saying that AI data centers are not a source of jobs, but a source of dollars. This is despite Argentine politicians, including Milei and others, claiming…
In case you need to see it to believe it, here is OpenAI's partner in Stargate Argentina saying that AI data centers are not a source of jobs, but a source of dollars. This is despite Argentine politicians, including Milei and others, claiming…
OpenVibe
open-source desktop environment for agentic coding, built with Tauri and Rust. In my testing, it uses up to 4.5× less peak RAM than comparable Electron-based tools.
The agent works directly with a real repository: it gathers context, creates a plan, edits files, runs commands, and shows its progress and tool calls directly in the chat.
The application combines Monaco Editor, reviewable diffs, change rollback, Git, a terminal, LSP, MCP integrations, and a dedicated Chromium instance that the agent can control through CDP.
OpenVibe runs locally and is not tied to a single AI provider. Users can connect a cloud API, a custom OpenAI-compatible endpoint, or a local model through Ollama, LM Studio, or vLLM.
Builds are available for Windows, macOS, and Linux. The project is licensed under GPL-3.0.
GitHub:
https://github.com/nihmadev/OpenVibe
#ai
open-source desktop environment for agentic coding, built with Tauri and Rust. In my testing, it uses up to 4.5× less peak RAM than comparable Electron-based tools.
The agent works directly with a real repository: it gathers context, creates a plan, edits files, runs commands, and shows its progress and tool calls directly in the chat.
The application combines Monaco Editor, reviewable diffs, change rollback, Git, a terminal, LSP, MCP integrations, and a dedicated Chromium instance that the agent can control through CDP.
OpenVibe runs locally and is not tied to a single AI provider. Users can connect a cloud API, a custom OpenAI-compatible endpoint, or a local model through Ollama, LM Studio, or vLLM.
Builds are available for Windows, macOS, and Linux. The project is licensed under GPL-3.0.
GitHub:
https://github.com/nihmadev/OpenVibe
#ai
GitHub
GitHub - nihmadev/OpenVibe
Contribute to nihmadev/OpenVibe development by creating an account on GitHub.