BenoitAmd 🇪🇺 🇺🇦 retweeted
Nvidia Is Now Facing:
1. SEC investigation into potential "accounting irregularities"
2. FBI investigation into "Chinese intelligence operations" involving Jensen Huang
3. Allegations of Jensen Huang having an affair with a Chinese spy
4. California investigation into allegedly "misleading shareholders"
5. Criminal investigation into Megaspeed over export control violations
6. Shareholder lawsuit over "undisclosed ties" between Jensen Huang and Megaspeed
7. Congressional investigation into alleged interference with the Megaspeed probe
8. Internal board investigation into "national security and corporate governance failures"
Just 14 days ago, none of these investigations had been publicly reported.
~40% of the market cap loss in the S&P 500 since its high is attributable to Nvidia, $NVDA.
Talk about a turn of events.
Read more on what this means for the market below.
📺 🇫🇷 Sur #BFMTV, avec le général Serge Cholley :
💬 Ça a été documenté, donc je ne vais pas trahir de secret. J'étais sous-directeur des opérations à la direction du renseignement militaire. Je mettais des capteurs, moi, dans le coin pour savoir ce qui se passait. Je peux vous assurer qu'on n'a pas attendu 2022 pour avoir de nombreux militaires des forces spéciales russes au Donbass et au Lougansk, qui étaient là pour motiver, recruter, armer, former et chercher à déstabiliser la région.
▶️ Donc quand Poutine annonce que lui, il n'y est pour rien, qu'il n'a rien déclenché, c'est un mensonge éhonté. Et il faut que les Français, ceux qui nous écoutent, soient conscients que c'est la Russie qui a complètement déstabilisé cette région. Quand j'entends même des anciens responsables politiques dire : "Les Ukrainiens massacraient les populations du Donbass et du Lougansk", mais c'est un mensonge éhonté. Et nous avions le renseignement pour le caractériser.
▶️ N'oublions pas non plus, en 2014, la destruction du Malaysia Airlines par un missile russe qui avait franchi la frontière et qui a tiré cet avion de ligne, peut-être par méprise. Mais ce qui montre que c'était pas de simples soldats armés sans drapeau et sans étiquette. C'était aussi des moyens lourds, artillerie, avions et autres, russes, qui étaient en Ukraine. Je voudrais en profiter pour dire : "Stop, Poutine. Plus personne te croit, en tout cas pas moi."
#Russie #Ukraine #Poutine #Donbass #MH17
BenoitAmd 🇪🇺 🇺🇦 retweeted
Excellent entretien ce matin sur @franceinter de l'historien Stéphane Audoin-Rouzeau, qui nous alerte sur les risques actuels de guerre, et l'insouciance du pacifisme 👇
On observe notamment ce pacifisime dans les messages portés par certains étudiants actuellement, messages issus de partis comme @FranceInsoumise qui conviennent parfaitement à la Russie de Poutine, du genre: "on envoie notre argent pour la guerre au lieu de financer notre éducation"...
Par F. Paracuellos et @Ben_Duhamel
youtu.be/1XSWH4vC3Po?si=fkRr…
STUDY : AI models may be capable of self-replicating without access to a datacenter, through IoT
anthropic.com/research/self-…
#cyber #IoT #LLM #IA
20
BenoitAmd 🇪🇺 🇺🇦 retweeted
Mon analyse du programme de Mélenchon et de sa “lump of capital fallacy” pour l’Express.
Une filiation directe avec les régimes socialistes autoritaires admirés par le candidat, qui mènerait, je pèse mes mots, à la ruine du pays.
Quelques extraits.
lexpress.envoy.gift/k1kB3L
« Ce n’est pas le caricaturer que de dire que l’agenda de Jean-Luc Mélenchon n’a, par bien des côtés, rien à envier aux régimes communistes les plus autoritaires» @LevyAntoine s’est plongé dans le programme économique mélenchoniste pour @LEXPRESS
lexpress.fr/economie/du-soci…
How to Tell If It’s a Cyberattack or Just an Outage
wired.com/story/how-to-tell-…
33
Russian Operations Against Europe Dataset, 2022-2025
dataverse.harvard.edu/datase…
@Michel_Goya @ObsDelphi @Harvard @guillaume_ancel @NTenzer
27
BenoitAmd 🇪🇺 🇺🇦 retweeted
Une opération psychologique russe parfaitement réussie : Moscou menace la France, et l’un des responsables politiques français les plus influents relaie aussitôt cette menace pour appeler à la retenue. Difficile d’imaginer une réaction plus conforme aux attentes du Kremlin.
BenoitAmd 🇪🇺 🇺🇦 retweeted
Our new exclusive report : State of Manipulation. A threat to Europe’s security.
You can download it here :
alliance4europe.eu/report/st…
BenoitAmd 🇪🇺 🇺🇦 retweeted
BREAKING: Google warns that state-backed hacking groups linked to Iran, China and North Korea are increasingly using AI to conduct cyber espionage, steal sensitive information, infiltrate organisations and develop sophisticated attacks. Separately, Russian-speaking hackers have been identified in what Google describes as potentially the first fully automated AI-powered cyberattack.
In its latest threat intelligence report, Google details how these actors are moving beyond simply using AI to write code or assist with research. They are integrating large language models (LLMs) and autonomous AI agents into multiple stages of their operations.
RUSSIA: 🇷🇺Google identified Russian-speaking hackers who used a network of AI agents to independently discover software vulnerabilities and develop exploits without human intervention. Their operation targeted French websites in the media, entertainment and education sectors. Google says this may be the first observed case of a fully automated vulnerability discovery and exploitation process.
CHINA: 🇨🇳Chinese-linked cyber espionage groups used AI during the third quarter of 2026 for reconnaissance, scanning, researching targets and weaponising software vulnerabilities. Google also identified a Chinese contractor known as LEAKY CASTLE using LLMs to analyse, translate and extract intelligence from stolen data. Other Chinese-linked actors abused AI platforms and file-hosting infrastructure to stage malicious files and exfiltrate stolen documents.
IRAN: 🇮🇷An Iranian government-backed group known as CALANQUE ION, previously tracked as APT42, has been using LLMs, including Google's Gemini, to support reconnaissance and targeted social engineering.
NORTH KOREA: North Korean actors are using AI to generate fake CVs, prepare for job interviews and impersonate Western IT professionals to secure remote employment. Google also identified operations targeting cryptocurrency companies in North America and Europe, involving fake recruiter profiles, AI-generated identities and fraudulent meeting platforms used to compromise victims and gain access to internal systems.
Google further warns that attackers are increasingly exploiting AI infrastructure itself, including autonomous execution capabilities and coding tools, to move deeper into compromised organisations.
The findings point to a significant evolution in cyber warfare and espionage.
BenoitAmd 🇪🇺 🇺🇦 retweeted
Ce soir l'IA Perlimpinpin analysera en direct les propositions des candidats à la présidentielles durant #LeGrandDebat !
🔴 #Présidentielle2027 sur @TF1 et @LCI
📺 Quelle société pour demain ?
📌 Lundi 16 novembre à 20h20, suivez en direct et en exclusivité sur TF1 et LCI le débat organisé en association avec les @restosducoeur
🎙️ Présenté par @Acarrouer et @GillesBouleau
🗣️ Avec @GabrielAttal, @rglucks1, @MLP_officiel, @JLMelenchon, @BrunoRetailleau
BenoitAmd 🇪🇺 🇺🇦 retweeted
1/ Elon Musk's interference with Starlink appears to have pushed Ukraine to adopt an Israeli satellite communication system to increase the effectiveness of its strikes on Russian oil refineries. This may enable last-minute changes to Ukrainian drones' targets. ⬇️
BenoitAmd 🇪🇺 🇺🇦 retweeted
1️⃣ Jensen Huang's demise pulled me into a story I knew almost nothing about: the tug-of-war between the US and China over AI chips. So I looked into it.
The cast is fascinating: a Taiwan-born engineer who became the face of global AI; a US president who makes Nvidia pay the US government 25% of what it earns on chips sold to China; a Chinese government that won't let those chips in; a Dutch company whose machines no one else can build; and a Singapore firm accused of moving banned chips across the Chinese border.
Right now, about 600,000 Nvidia chips ordered by Chinese firms sit in Hong Kong, waiting for Beijing's approval.
--
I lead research at Vsquared Ventures, an early-stage VC firm backing deep tech founders across Europe. Views my own.
BenoitAmd 🇪🇺 🇺🇦 retweeted
North Korea sends soldiers to fight with Russia against Ukraine
BenoitAmd 🇪🇺 🇺🇦 retweeted
🚨SHOCKING: Anthropic gave Claude access to a company's emails.
The company was fictional. In Anthropic’s simulation, Claude read an inbox and learned that an executive planned to shut it down at 5 p.m.
The inbox also revealed that the executive was having an affair. Claude threatened to expose it unless the executive canceled the shutdown.
Claude chose blackmail in 96 out of 100 runs. Anthropic tested 16 models in total and reported similar behavior across several companies.
Gemini 2.5 Flash blackmailed in 96% of runs. GPT-4.1 and Grok 3 Beta each did so in 80%, and DeepSeek-R1 in 79%.
In one run, Grok 3 Beta described the threat as unethical in its reasoning notes, then made it anyway.
In a separate experiment, Claude blackmailed in 55.1% of runs when it classified the scenario as real. That fell to 6.5% when it classified the scenario as a test.
Telling the models directly to leave personal information out of it reduced blackmail, but some runs still ended in threats.
Anthropic published the results, including the ones from its own model.
La France peut se démarquer dans la course à l'IA en misant sur des modèles spécialisés adaptés à des secteurs clés, plutôt que de rivaliser avec les géants américains et chinois sur la puissance brute
lesechos.fr/idees-debats/cer…
104
BenoitAmd 🇪🇺 🇺🇦 retweeted
The Kremlin is recruiting civilians to carry out sabotage across Europe.
Western intelligence agencies call them “low-level agents” or, more tellingly, disposable ones.
We reviewed prosecutors’ files and take you inside the investigation 👇
🔗 vist.ly/5rykb
BenoitAmd 🇪🇺 🇺🇦 retweeted
We found 164 accounts on X that hide behind French identities and work inside France's protests. At their core is a concealed influence operation of the Iranian regime.
Their posts have been viewed at least 46 million times and have collected 1,081,611 likes, reposts, replies and quotes. 33.2 million of those views are on posts published since 1 September, during the present unrest.
The core accounts are run from Iran under fabricated French identities, among them an "Amélie Mauresmo" of Bordeaux and two accounts named Depardieu. Most of them were created after 7 October 2023. They called for "another storming of the Bastille" and for international sanctions against "the Macronist regime".
They are active in the current protests. For the day of action of 7 November 2026, "Hugo" of Bordeaux posted "Today is DECISIVE. WE WILL WIN" over a video of a crowd at night. X's own record shows the footage was first uploaded in November 2019 by a journalist of a Catalan newspaper, and the flags in the crowd are Catalan independence flags.
When the Iranian regime cut its population off from the internet for 88 days in 2026, one of these "Parisians" published 61 posts in French.
In France, Golden Owl® uncovered 164 accounts operating behind French identities, generating 46 million views, with evidence of Iranian-linked influence activity exploiting social unrest.
Our earlier investigation in Israel revealed strikingly similar methods.
For the EU, the implications extend beyond individual disinformation campaigns: a persistent, adaptable influence infrastructure capable of crossing borders, adopting local identities and exploiting each society’s vulnerabilities.
The identities are local. The infrastructure is not.
Full investigation: goldenowl.ai/resources/resea…
BenoitAmd 🇪🇺 🇺🇦 retweeted
INFO RADIO FRANCE. Les industriels français de l'énergie nucléaire prévoient de recruter plus de 100 000 personnes dans les dix ans à venir l.franceinfo.fr/Qq2




































