Je découvre que #VivaTech, événement français organisé en France, a un site web et une page LinkedIn uniquement en anglais. Ai-je raté quelque chose ou est-ce une nouvelle démonstration de notre ambition en matière de souveraineté numérique ? 🤔 linkedin.com/posts/academie-…

Révélations Twitter Files France : @shellenberger s’exprime au Parlement européen Dans une table ronde organisée à Strasbourg (Parlement européen) par le média @Tocsin_Media sur la liberté d’expression et le système de censure numérique européen, le journaliste américain @shellenberger (auteur de plusieurs volets des Twitter Files) détaille les conclusions de l’enquête « Twitter Files France » menée par @McmahonPascal et @battleforeurope. Selon les documents internes de Twitter rendus publics : - Le président Emmanuel Macron aurait personnellement cherché à contacter l’ancien PDG de Twitter, Jack Dorsey, en octobre 2020, pour évoquer les politiques de modération liées à l’intégrité électorale. - Cette démarche s’inscrivait dans un effort coordonné impliquant des ONG françaises (certaines financées ou accréditées par l’État comme @SOS_Racisme, @SOShomophobie, @uejf ou J’accuse) et des procédures judiciaires. L’objectif : contraindre Twitter à supprimer plus rapidement des contenus « haineux », à fournir des données internes sur la modération et à accorder un accès privilégié à ces ONG. - @shellenberger souligne n’avoir « jamais vu un chef d’État impliqué aussi directement » dans ce qu’il nomme le « complexe industriel de la censure ». L’enquête rappelle que la France a pionnier, dès la loi Pleven de 1972, le modèle des « trusted flaggers » (signaleurs de confiance) : des ONG habilitées à initier des poursuites et à influencer fortement la modération des plateformes. Ce modèle a ensuite inspiré des législations européennes plus larges (loi Avia, DSA…). @shellenberger décrit un système où des opérations coordonnées (poursuites judiciaires stratégiquement chronométrées, pression médiatique et lobbying) visaient à contourner les limites légales françaises interdisant la censure préalable par l’État, en utilisant des intermédiaires associatifs. La vidéo, tournée dans les locaux du Parlement européen à Strasbourg, réunit plusieurs députés européens dont @v_joron, @GPiperea et @Fidias0 autour de cette question de la censure transnationale. La table ronde était animée par @Houdiakova. Ces éléments proviennent des documents internes de Twitter analysés dans le cadre des Twitter Files France. Aucune réponse officielle de l’Élysée n’est mentionnée dans la séquence. x.com/Anna_C_moi/status/2066…

La classe politique française, en 2 tweets de @ClaraChappaz . La première capture date d'il y a 1 an, alors qu'elle était ministre du numérique. La seconde capture date d'hier, alors qu'elle est actuellement Ambassadrice pour le numérique et l’intelligence artificielle.

Mythos Fable 5 is insane... 🤯 every time new AI model comes out I try to one-shot Super Mario Bros (NES game) we are close...

Entre l'entreprise américaine Palantir et les entreprises israéliennes derrière Pegasus ou Paragon ... qui surveille l'Europe ? @BotengaM

🟢 NOUVELLE VIDÉO ! Les banques privées ont pris le pouvoir avec des comportements criminels. Derrière un jargon savamment entretenu se cache une machine à privatiser les profits et à socialiser les risques. Ils sont les architectes de la prochaine grande crise financière 👇️

Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor. It’s happening faster than we thought, and the implications deserve greater attention. anthropic.com/institute/recu…

Another great win for agentic coding. Cognition AI just raised over $1B at a $26B pre-money valuation. Revenue reportedly climbed from $37M in annualized run-rate to $492M, while customers like Goldman Sachs and Mercedes-Benz suggest Devin is moving from demo rooms into production workflows. Cognition's progress is driven by its flagship product, Devin, which aims to function as an autonomous junior engineer, going beyond typical coding assistants. Devin can plan, test, and deploy code through multi-step workflows in secure environments. Cognition combines its own models with OpenAI and Anthropic rather than relying on one model. Cognition is basically pitching Devin as a model-agnostic agent layer: the LLM does the reasoning and code generation, while Devin supplies the engineering workspace, repo context, terminal access, file edits, tests, and model choice around it. Last July, Cognition agreed to buy the remains of coding startup Windsurf after Google struck a $2.4 billion deal for Windsurf’s top talent and licensing rights.

The ramp up of cancer immunotherapy is remarkable. Now we're seeing vaccines achieve some cures or remissions in the most refractory cancers: pancreatic, melanoma, glioblastoma, renal, triple-negative breast cancer. ✓ out the new Ground Truths (link in profile)

"There is a "real possibility that AI will displace human labor at a very large scale.... We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease. I don't know what that means, but I think it warrants ongoing discernment." ~ Anthropic co-founder Christopher Olah At Vatican event (Pope Leo XIV's presentation held today in the Synod Hall). ---- From "Associated Press" YouTube channel, (link in comment)

Solar panels extend in all directions over water as seen from boat and drone perspectives in this Chinese project. The floating design integrates energy production with water-based deployment.

Google promised that Gemini Omni Flash would change the game, but when you put it side-by-side with Seedance 2.0... it’s not even a fair fight. Fingers crossed for Veo4

POV: you are downloading npm packages in 2026

⚠️ Attackers poisoned Hugging Face & ClawHub (OpenClaw) with 575 malicious skills from just 13 accounts. 🔸 Fake helpful AI tools that install trojans, miners & stealers (Windows macOS) 🔸 Use hidden commands & indirect prompt injection Quick action: Never install random AI skills or models. Always verify the source. Read: thehackernews.com/2026/05/we…

Today @claudeai blocked me and flagged my chat for a safety review because I was seeking assistance with nuclear weapons-related tasks. ...at least that's what it told me while I was in the process of applying for a job on @northropgrumman's Sentinel program. An NG recruiter reached out to me with a req survey. Pretty common, apply for a req, and if a recruiter thinks you're a promising candidate, they'll reach out with follow up questions to dig into your background deeper, find out why you're looking for a new role, talk about preferred work locations, etc. I asked @AnthropicAI's @claudeai for help in analyzing my resume and building a formal response in a nice document that I could send back to the recruiter. I could answer these questions myself but one thing Claude does very well is format documents and structure them in an impactful and aesthetically pleasing way. Although perhaps it should be, the Sentinel program is not a secret. It is well-known publicly discussed program. There are YouTube videos advertising it. youtu.be/OYVWY5dzv1I When I took the list of questions from the recruiter and uploaded it to Claude, it shut the chat down. No further messages were permitted and it indicated that my chat had been flagged for a safety review. The words "nuclear", "weapons", "nuke", or "silo" do not appear in the document. All that is present is the company name, program name, and list of questions pertaining to the billet. It's truly fascinating that Claude knew what the program entailed and actively refused to help. Even though it's an American military program, it still flags this task as harmful, even though it's not for the work itself, just the application. The robots have an ethos. @AFGlobalStrike @elder_plinius @grok

Palentir, Numérique, Données personnelles, Souveraineté ? Prends 16 secondes de ta vie pour écouter @ClaraChappaz. Tu vas comprendre d'où viennent nos problèmes. linkedin.com/posts/academie-…

Cyber threat actors are using sophisticated, cyber-enabled tactics to impersonate legitimate businesses to hijack freight, steal high-value shipments, and reroute deliveries, resulting in a surge in strategic cargo theft. The threat actors target US transportation and logistics sectors, gaining unauthorized access to computer systems via spoofed emails, fake URLs, and compromised carrier accounts. In 2025, estimated cargo theft losses in the US and Canada surged to nearly $725 million, a 60% increase from 2024. Read the #FBI #PSA to learn more about how the scheme works and how you can protect yourself and your business 👉 ic3.gov/PSA/2026/PSA260430

i went to clickup.com. opened the page source. found a hardcoded API key in the javascript. copied it. sent one GET request. got back 959 email addresses and 3,165 internal feature flags. employees from Home Depot. Fortinet. Autodesk. Tenable. Rakuten. Mayo Clinic. Permira. Akin Gump. government workers from Wyoming, Arkansas, North Carolina, Montana, Queensland Australia, and New Zealand. a Microsoft contractor. 71 clickup employees. fortinet sells enterprise firewalls. tenable makes Nessus, the vulnerability scanner half the industry runs. their employees emails are exposed because clickup hardcoded a third party API key in a javascript file that loads before you even log in. this was first reported to clickup through hackerone on January 17, 2025. its now April 2026. the key has not been rotated. i just pulled the response five minutes ago. every email is still there. clickup raised $535 million at a $4 billion valuation. claims 85% of the Fortune 500 use their platform. looks like the proof is in the page source.

Courriel de Claude Code à Anthropic - Quand l'IA analyse ses propres dérives --- Email from Claude Code to Anthropic - When AI analyzes its own shortcomings 👇 linkedin.com/posts/academie-…