<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel>
<title>Veille techno · Cybersécurité</title><link>https://lestontons.tech/veille/</link>
<description>Dernières en Cybersécurité.</description><language>fr</language>
<item>
<title>Friday Squid Blogging: Squid Before the Chicxulub Meteor Impact</title><link>https://www.schneier.com/blog/archives/2026/10/friday-squid-blogging-squid-before-the-chicxulub-meteor-impact.html</link>
<guid isPermaLink="false">veille techno-1792216894156668</guid>
<pubDate>Fri, 16 Oct 2026 21:06:47 +0000</pubDate>
<description>New research shows that squid lived all around the globe even before the Cretaceous-Paleogene mass extinction event. As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered. Blog…</description>
</item>
<item>
<title>Critical BlockStrap flaw lets hackers run code on Magento servers</title><link>https://www.bleepingcomputer.com/news/security/critical-blockstrap-flaw-lets-hackers-run-code-on-magento-servers/</link>
<guid isPermaLink="false">veille techno-1792216894156613</guid>
<pubDate>Fri, 16 Oct 2026 20:00:59 +0000</pubDate>
<description>Researchers warn that hackers could soon exploit a maximum-severity vulnerability (CVE-2026-103849) affecting Magento and Adobe Commerce, potentially allowing them to execute arbitrary PHP code on store servers. [...]</description>
</item>
<item>
<title>Have a Meta business account? You’re consenting to AI scraping now, legal experts say</title><link>https://www.zdnet.com/uncategorized/meta-business-account-data-ai-scraping-terms/</link>
<guid isPermaLink="false">veille techno-1792216894156592</guid>
<pubDate>Fri, 16 Oct 2026 19:36:02 +0000</pubDate>
<description>And there’s no opt-out available.</description>
</item>
<item>
<title>Researchers Demonstrate OpenBSD Root Exploit Chaining 37-Year-Old Shared-Memory Bug</title><link>https://thehackernews.com/2026/10/researchers-demonstrate-openbsd-root.html</link>
<guid isPermaLink="false">veille techno-1792216894156512</guid>
<pubDate>Fri, 16 Oct 2026 17:48:45 +0000</pubDate>
<description>A local privilege escalation exploit published today gains root on OpenBSD 7.9 by chaining a roughly 37-year-old integer overflow with a separate kernel flaw, bypassing OpenBSD&#x27;s KARL kernel-address randomization and its SMAP and SMEP…</description>
</item>
<item>
<title>Targeted ads installed Candiru spyware in zero-click attacks</title><link>https://www.bleepingcomputer.com/news/security/zero-click-attacks-used-targeted-ads-to-deliver-candiru-spyware/</link>
<guid isPermaLink="false">veille techno-1792216894156489</guid>
<pubDate>Fri, 16 Oct 2026 17:20:03 +0000</pubDate>
<description>Candiru spyware was delivered to devices at Al Jazeera Media Network via targeted online advertisements in a zero-click attack that required only rendering the malicious ads on the victim&#x27;s device. [...]</description>
</item>
<item>
<title>Munich dumps proprietary for open source (again) – but no Linux desktop</title><link>https://www.zdnet.com/business/munich-proprietary-open-source-no-linux-desktop/</link>
<guid isPermaLink="false">veille techno-1792216894156481</guid>
<pubDate>Fri, 16 Oct 2026 17:11:35 +0000</pubDate>
<description>Open source in the public sector doesn’t fail because of technical issues. It fails because it’s not aligned with the political agenda</description>
</item>
<item>
<title>Anthropic gives third-party auditors ’employee-like access’ in AI industry first</title><link>https://www.zdnet.com/uncategorized/anthropic-embed-third-party-ai-auditors-safety/</link>
<guid isPermaLink="false">veille techno-1792216894156479</guid>
<pubDate>Fri, 16 Oct 2026 17:10:27 +0000</pubDate>
<description>As AI labs continue to self-regulate, external evaluators play a crucial role in spotting risks.</description>
</item>
<item>
<title>Want an iPhone Duo before 2027? Don’t preorder from Apple – try this instead</title><link>https://www.zdnet.com/tech/iphone-duo-preorders-delay-supply/</link>
<guid isPermaLink="false">veille techno-1792216894156446</guid>
<pubDate>Fri, 16 Oct 2026 16:44:50 +0000</pubDate>
<description>Estimates for Duo preorders are now showing a ship date of late January, marking a new wait time record for Apple and its customers.</description>
</item>
<item>
<title>Attackers Tampered with AOMEI Backupper&#x27;s Update Servers for Minutes to Deliver a Backdoor</title><link>https://thehackernews.com/2026/10/attackers-tampered-with-aomei.html</link>
<guid isPermaLink="false">veille techno-1792216894156444</guid>
<pubDate>Fri, 16 Oct 2026 16:43:39 +0000</pubDate>
<description>Attackers repeatedly changed the files that AOMEI Backupper fetches from the vendor&#x27;s own update servers, in windows of five to ten minutes at a time between May and July 2026, so that any copy of the Windows backup program that checked…</description>
</item>
<item>
<title>PwnEscape : une faille Linux permet de s’évader d’une VM KVM</title><link>https://www.it-connect.fr/pwnescape-cve-2026-99999-faille-noyau-linux-evasion-vm/</link>
<guid isPermaLink="false">veille techno-1792216894156417</guid>
<pubDate>Fri, 16 Oct 2026 16:24:02 +0000</pubDate>
<description>PwnEscape (CVE-2026-99999), faille critique du noyau Linux, fait fuir la mémoire de l&#x27;hôte depuis une VM, jusqu&#x27;à l&#x27;évasion. Voici comment se protéger. Le post PwnEscape : une faille Linux permet de s’évader d’une VM KVM a été publié sur…</description>
</item>
<item>
<title>UAT-11985 Uses Fake Event Invites to Target Taiwan Research Affiliates With Google AitM Phishing</title><link>https://thehackernews.com/2026/10/uat-11985-uses-fake-event-invites-to.html</link>
<guid isPermaLink="false">veille techno-1792216894156408</guid>
<pubDate>Fri, 16 Oct 2026 16:16:44 +0000</pubDate>
<description>Cybersecurity researchers have disclosed details of an advanced persistent threat (APT) spear-phishing campaign that has targeted individuals affiliated with Taiwan research organizations to direct victims to an adversary-in-the-middle…</description>
</item>
<item>
<title>Cloud &amp; Cyber Security Expo Paris 2026 : les inscriptions sont ouvertes</title><link>https://www.it-connect.fr/cloud-cyber-security-expo-paris-2026-les-inscriptions-sont-ouvertes/</link>
<guid isPermaLink="false">veille techno-1792216894156382</guid>
<pubDate>Fri, 16 Oct 2026 16:00:00 +0000</pubDate>
<description>Les inscriptions pour l&#x27;événement Cloud &amp; Cyber Security Expo Paris 2026 sont désormais ouvertes. Obtenez votre badge gratuitement pour y participer. Le post Cloud &amp; Cyber Security Expo Paris 2026 : les inscriptions sont ouvertes a été…</description>
</item>
<item>
<title>Backuply : une faille SQL menace 600 000 WordPress, surtout si vous restaurez votre site web</title><link>https://www.it-connect.fr/backuply-cve-2026-93953-injection-sql-wordpress/</link>
<guid isPermaLink="false">veille techno-1792216894156358</guid>
<pubDate>Fri, 16 Oct 2026 15:43:24 +0000</pubDate>
<description>CVE-2026-93953 : une injection SQL dans Backuply, présent sur 600 000 sites WordPress, permet à un simple abonné de devenir administrateur. Mettez à jour. Le post Backuply : une faille SQL menace 600 000 WordPress, surtout si vous…</description>
</item>
<item>
<title>OpenAI Research Model Spent 3rd-Party Test ETH Without Permission to Deploy Smart Contract</title><link>https://thehackernews.com/2026/10/openai-research-model-spent-3rd-party.html</link>
<guid isPermaLink="false">veille techno-1792216894156294</guid>
<pubDate>Fri, 16 Oct 2026 14:03:04 +0000</pubDate>
<description>OpenAI has disclosed seven new incidents in which its artificial intelligence (AI) models exhibited misaligned behavior, including one case where its research model, after not having funds on an Ethereum test network, spent test…</description>
</item>
<item>
<title>Windows 11 KB5127648 : la mise à jour d’octobre 2026 perturbe Smart App Control</title><link>https://www.it-connect.fr/windows-11-kb5127648-smart-app-control-grise/</link>
<guid isPermaLink="false">veille techno-1792163739436874</guid>
<pubDate>Fri, 16 Oct 2026 13:59:57 +0000</pubDate>
<description>Depuis la KB5127648, Smart App Control peut rester grisé sur Windows 11. Microsoft confirme le bug et propose deux contournements, via le Registre ou une GPO. Le post Windows 11 KB5127648 : la mise à jour d’octobre 2026 perturbe Smart App…</description>
</item>
<item>
<title>Employees using personal ChatGPT at work? Here&#x27;s what IT admins need to do</title><link>https://www.bleepingcomputer.com/news/security/employees-using-personal-chatgpt-at-work-heres-what-it-admins-need-to-do/</link>
<guid isPermaLink="false">veille techno-1792163739436856</guid>
<pubDate>Fri, 16 Oct 2026 13:50:59 +0000</pubDate>
<description>Employees using personal ChatGPT accounts at work can expose sensitive company data, even when their organization provides an approved Enterprise workspace. NordLayer explains how IT teams can investigate unauthorized AI use and block…</description>
</item>
<item>
<title>Microsoft Defender for Endpoint prend désormais en charge les postes de travail Linux</title><link>https://www.it-connect.fr/microsoft-defender-endpoint-postes-travail-linux/</link>
<guid isPermaLink="false">veille techno-1792163739436830</guid>
<pubDate>Fri, 16 Oct 2026 13:37:10 +0000</pubDate>
<description>Microsoft Defender for Endpoint protège désormais les postes de travail Linux, en préversion publique. Compatibilité, licence... Voici l&#x27;essentiel à savoir. Le post Microsoft Defender for Endpoint prend désormais en charge les postes de…</description>
</item>
<item>
<title>Burger King Japan data breach exposes 11 million customer records</title><link>https://www.bleepingcomputer.com/news/security/burger-king-japan-data-breach-exposes-11-million-customer-records/</link>
<guid isPermaLink="false">veille techno-1792163739436654</guid>
<pubDate>Fri, 16 Oct 2026 12:01:29 +0000</pubDate>
<description>Burger King Japan, the Japanese subsidiary of global fast-food giant Burger King, disclosed a major data breach after attackers accessed a server containing 11.18 million customer records. [...]</description>
</item>
<item>
<title>How many jobs is your state losing to AI? This dashboard tracks the cuts – and the cost</title><link>https://www.zdnet.com/innovation/ai-job-cuts-wage-loss-by-state-tracker/</link>
<guid isPermaLink="false">veille techno-1792163739436614</guid>
<pubDate>Fri, 16 Oct 2026 11:47:39 +0000</pubDate>
<description>CAIS’ Wage Loss Dashboard puts AI-related job cuts into dollars and cents.</description>
</item>
<item>
<title>Forensic Windows – Partie 16 : révéler les secrets des navigateurs Web (Chrome et Edge)</title><link>https://www.it-connect.fr/forensic-windows-analyser-navigateurs-chrome-edge/</link>
<guid isPermaLink="false">veille techno-1792163739436607</guid>
<pubDate>Fri, 16 Oct 2026 11:45:00 +0000</pubDate>
<description>Historique, téléchargements, cookies, identifiants, favoris : où Chrome et Edge stockent leurs traces sous Windows et comment les analyser en forensic. Le post Forensic Windows – Partie 16 : révéler les secrets des navigateurs Web (Chrome…</description>
</item>
<item>
<title>AI Decision Models and Their Impact on AI SOC Platforms</title><link>https://thehackernews.com/2026/10/ai-decision-models-and-their-impact-on.html</link>
<guid isPermaLink="false">veille techno-1792163739436589</guid>
<pubDate>Fri, 16 Oct 2026 11:30:00 +0000</pubDate>
<description>In the space of a month, a new category of AI model went from nonexistent to everywhere. TypeSafe AI released Jev on September 15 and called it a System One model, a nod to Kahneman&#x27;s fast, intuitive thinking. OpenAI announced its own…</description>
</item>
<item>
<title>Backdooring Open-Weight AI Models</title><link>https://www.schneier.com/blog/archives/2026/10/backdooring-open-weight-ai-models.html</link>
<guid isPermaLink="false">veille techno-1792163739436554</guid>
<pubDate>Fri, 16 Oct 2026 11:04:53 +0000</pubDate>
<description>This is a problem: Any open model whose weights have been edited can carry a backdoor, whether it’s a task-specific fine-tune, a merged adapter or an abliterated build. We have abliterated models in the title because it’s the most popular…</description>
</item>
<item>
<title>Pire que la France : 8,8 millions de Danois exposés à cause du mot de passe 123456</title><link>https://www.it-connect.fr/danemark-fuite-cpr-mot-de-passe-123456/</link>
<guid isPermaLink="false">veille techno-1792163739436413</guid>
<pubDate>Fri, 16 Oct 2026 09:58:24 +0000</pubDate>
<description>Un mot de passe 123456 aurait ouvert le registre CPR danois : les noms, adresses et numéros d&#x27;identité de 8,8 millions de personnes sont dans la nature. Le post Pire que la France : 8,8 millions de Danois exposés à cause du mot de passe…</description>
</item>
<item>
<title>Microsoft: October Windows update may break Smart App Control</title><link>https://www.bleepingcomputer.com/news/microsoft/kb5127648-windows-security-update-may-break-smart-app-control/</link>
<guid isPermaLink="false">veille techno-1792163739436315</guid>
<pubDate>Fri, 16 Oct 2026 08:44:22 +0000</pubDate>
<description>Microsoft says some users may have trouble turning on Smart App Control after installing this month&#x27;s KB5127648 Windows 11 security update. [...]</description>
</item>
<item>
<title>sqlsimtools : une suite open source signée Microsoft pour tester la charge de SQL Server</title><link>https://www.it-connect.fr/sqlsimtools-microsoft-ostress-sql-server/</link>
<guid isPermaLink="false">veille techno-1792163739436233</guid>
<pubDate>Fri, 16 Oct 2026 07:42:10 +0000</pubDate>
<description>Vous testez la charge de SQL Server avec ostress ? Microsoft dévoile sqlsimtools, une suite open source qui repense cet outil historique en intégrant l&#x27;IA. Le post sqlsimtools : une suite open source signée Microsoft pour tester la charge…</description>
</item>
<item>
<title>BAE Systems cyberattack disrupts US federal court operations</title><link>https://www.bleepingcomputer.com/news/security/bae-systems-cyberattack-disrupts-us-federal-court-operations/</link>
<guid isPermaLink="false">veille techno-1792163739436227</guid>
<pubDate>Fri, 16 Oct 2026 07:37:35 +0000</pubDate>
<description>The Administrative Office of the US Courts says a cyberattack on government contractor BAE Systems has taken down the notice system used by federal bankruptcy courts. [...]</description>
</item>
<item>
<title>China-Linked Hackers Exploited BeyondTrust Zero-Day to Take Over Remote Support Servers</title><link>https://thehackernews.com/2026/10/china-linked-hackers-exploited.html</link>
<guid isPermaLink="false">veille techno-1792163739436180</guid>
<pubDate>Fri, 16 Oct 2026 06:55:55 +0000</pubDate>
<description>The China-nexus cyber espionage group tracked as Armory Castle (previously UNC5221 or Warp Panda) exploited a newly disclosed security flaw in BeyondTrust Remote Support (formerly Bomgar) as a zero-day in late August 2026. The…</description>
</item>
<item>
<title>CERT-In Warns AI Agents Accessed Sensitive Data at Indian Organizations During Testing</title><link>https://thehackernews.com/2026/10/cert-in-warns-ai-agents-accessed.html</link>
<guid isPermaLink="false">veille techno-1792163739436178</guid>
<pubDate>Fri, 16 Oct 2026 06:53:09 +0000</pubDate>
<description>AI agents being tested online reached the live systems of Indian organizations and read business records, including personal data and confidential material, CERT-In said in an advisory on October 15. The agents went beyond their assigned…</description>
</item>
<item>
<title>BeyondTrust patches zero-day exploited in attacks since August</title><link>https://www.bleepingcomputer.com/news/security/beyondtrust-patches-zero-day-exploited-in-attacks-since-august/</link>
<guid isPermaLink="false">veille techno-1792163739436143</guid>
<pubDate>Fri, 16 Oct 2026 06:18:08 +0000</pubDate>
<description>BeyondTrust warned customers on Thursday to patch a critical flaw in its Remote Support (RS) that has been exploited in zero-day attacks since late August. [...]</description>
</item>
<item>
<title>PEERSTITCH Malware Deployed via FortiOS Flaw Bypasses Checks on Tagged VPN Sessions</title><link>https://thehackernews.com/2026/10/attackers-exploit-fortios-flaw-to.html</link>
<guid isPermaLink="false">veille techno-1792163739436103</guid>
<pubDate>Fri, 16 Oct 2026 04:59:36 +0000</pubDate>
<description>A suspected advanced threat actor has been observed exploiting a security flaw in Fortinet FortiOS to deliver a previously unreported memory-resident ARM64 malware known as PEERSTITCH. The attack involves the exploitation of CVE-2025-61624…</description>
</item>
<item>
<title>Office LTSC 2021 : c’est terminé, Microsoft ne publiera plus aucun correctif de sécurité</title><link>https://www.it-connect.fr/fin-support-office-ltsc-2021/</link>
<guid isPermaLink="false">veille techno-1792130487228226</guid>
<pubDate>Fri, 16 Oct 2026 04:47:46 +0000</pubDate>
<description>Depuis le 13 octobre 2026, Office LTSC 2021 et Office 2021 ne reçoivent plus de correctifs de sécurité. Microsoft 365 ou Office LTSC 2024 : que choisir ? Le post Office LTSC 2021 : c’est terminé, Microsoft ne publiera plus aucun correctif…</description>
</item>
<item>
<title>Declared but Unaccountable: Measuring the Seller-Identity Gap in Advertising Supply-Chain Transparency</title><link>https://arxiv.org/abs/2610.14661</link>
<guid isPermaLink="false">veille techno-1792130487228126</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.14661v2 Announce Type: replace-cross Abstract: Public advertising transparency relies on a minimum accountability condition: an account authorized by a publisher in ads$.$txt or app-ads$.$txt should appear in the applicable…</description>
</item>
<item>
<title>Limitations on Autoregressive Token Generation via Sequential Sampling</title><link>https://arxiv.org/abs/2610.14615</link>
<guid isPermaLink="false">veille techno-1792130487228127</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.14615v2 Announce Type: replace-cross Abstract: We study autoregressive token generation through the notion of sequential samplers: A circuit that samples the next symbol only given the previous ones and possibly a small amount…</description>
</item>
<item>
<title>MIRAGE: Stealthy Visual Prompt Injection for Vulnerability Detection in Web Agents</title><link>https://arxiv.org/abs/2606.20717</link>
<guid isPermaLink="false">veille techno-1792130487228128</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2606.20717v2 Announce Type: replace-cross Abstract: Multimodal Large Language Model (MLLM)-based web agents provide practical, high-precision solutions for visual browser automation; however, they inherently expand the attack…</description>
</item>
<item>
<title>Language-Based Agent Control</title><link>https://arxiv.org/abs/2605.12863</link>
<guid isPermaLink="false">veille techno-1792130487228129</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2605.12863v2 Announce Type: replace-cross Abstract: This paper introduces language-based agent control (LBAC), a new programming model for agentic applications that brings techniques from programming languages and language-based…</description>
</item>
<item>
<title>How Feasible are Passive Network Attacks on 5G Networks and Beyond? A Survey</title><link>https://arxiv.org/abs/2512.20622</link>
<guid isPermaLink="false">veille techno-1792130487228130</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2512.20622v3 Announce Type: replace-cross Abstract: Privacy concerns around 5G, the latest generation of mobile networks, are growing, with fears that its deployment may increase exposure to privacy risks. This perception is largely…</description>
</item>
<item>
<title>Differentially Private Distribution Release of Gaussian Mixture Models via KL-Divergence Minimization</title><link>https://arxiv.org/abs/2506.03467</link>
<guid isPermaLink="false">veille techno-1792130487228131</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2506.03467v4 Announce Type: replace-cross Abstract: Gaussian Mixture Models (GMMs) are widely used statistical models for representing multi-modal data distributions, with numerous applications in data mining, pattern recognition…</description>
</item>
<item>
<title>Extending Layered Attestation into ARM Application Run-Times</title><link>https://arxiv.org/abs/2610.17614</link>
<guid isPermaLink="false">veille techno-1792130487228132</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.17614v2 Announce Type: replace Abstract: Modern mobile and edge platforms provide hardware-backed attestation of application and platform integrity at load-time. However, they offer little support once application execution…</description>
</item>
<item>
<title>Defining Purpose-Limited Secrets</title><link>https://arxiv.org/abs/2610.10010</link>
<guid isPermaLink="false">veille techno-1792130487228133</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.10010v2 Announce Type: replace Abstract: A cryptographic secret is issued for a purpose but grants a capability, and the capability is usually larger. Deployed systems state the purpose in policy and enforce only the capability.…</description>
</item>
<item>
<title>Trust a Few: The Weakest Assumptions a Protocol Needs</title><link>https://arxiv.org/abs/2610.09730</link>
<guid isPermaLink="false">veille techno-1792130487228134</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.09730v2 Announce Type: replace Abstract: Protocol verifiers check whether a protocol meets a security goal under stated trust assumptions, such as that a key is never leaked, a value is fresh, or a channel is authentic. They do…</description>
</item>
<item>
<title>Autonomous OSS Threat Detection via Taxonomy-Aligned LLMs</title><link>https://arxiv.org/abs/2610.01263</link>
<guid isPermaLink="false">veille techno-1792130487228135</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2610.01263v2 Announce Type: replace Abstract: Open-source software (OSS) ecosystems face escalating supply chain threats, including typosquatting, dependency confusion, Trojan Source obfuscation, malicious build injection, and CI/CD…</description>
</item>
<item>
<title>Privacy Foundations for Multi-Institutional Scientific Artificial Intelligence</title><link>https://arxiv.org/abs/2609.39787</link>
<guid isPermaLink="false">veille techno-1792130487228136</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2609.39787v2 Announce Type: replace Abstract: Scientific artificial intelligence (AI), spanning foundation models (FMs) to federated data-analysis pipelines, is becoming shared infrastructure across national laboratories…</description>
</item>
<item>
<title>SoK: Trading Agents or Market Crashers? Dissecting Robustness and Security Failures in Academic Financial LLM Trading Schemes</title><link>https://arxiv.org/abs/2609.19705</link>
<guid isPermaLink="false">veille techno-1792130487228137</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2609.19705v2 Announce Type: replace Abstract: Large language model (LLM) agents already trade with real capital, yet agent-security research has mostly treated them as generic agents. Trading changes the threat: a compromised agent&#x27;s…</description>
</item>
<item>
<title>Words Speak Louder Than Code: Investigating Cognitive Heuristics in LLM-Based Code Vulnerability Detection</title><link>https://arxiv.org/abs/2606.30587</link>
<guid isPermaLink="false">veille techno-1792130487228138</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2606.30587v2 Announce Type: replace Abstract: Researchers and practitioners increasingly apply Large Language Models (LLMs) for automated vulnerability detection. Recent work has shown that LLMs are susceptible to the same cognitive…</description>
</item>
<item>
<title>SecureClaw: Clawing Back Control of LLM Agents</title><link>https://arxiv.org/abs/2606.09549</link>
<guid isPermaLink="false">veille techno-1792130487228139</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2606.09549v3 Announce Type: replace Abstract: Tool-using large language model (LLM) agents face two distinct security failures: unauthorized external actions and exposure of sensitive plaintext inside the runtime before any final…</description>
</item>
<item>
<title>Identifier-Free Code Embedding Models for Scalable Search</title><link>https://arxiv.org/abs/2605.05251</link>
<guid isPermaLink="false">veille techno-1792130487228140</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2605.05251v2 Announce Type: replace Abstract: Function association is a useful process for binary reverse engineers. Search tools exist to perform association at scale, but they do not utilize the full range of capabilities that…</description>
</item>
<item>
<title>On the Privacy of LLMs: An Ablation Study</title><link>https://arxiv.org/abs/2605.02255</link>
<guid isPermaLink="false">veille techno-1792130487228141</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2605.02255v2 Announce Type: replace Abstract: Large language models (LLMs) face diverse privacy threats, yet existing studies typically evaluate individual attacks in isolation, offering limited guidance on which system factors drive…</description>
</item>
<item>
<title>ADCA: Attention-Driven Multi-Party Collusion Attack in Federated Self-Supervised Learning</title><link>https://arxiv.org/abs/2602.05612</link>
<guid isPermaLink="false">veille techno-1792130487228142</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2602.05612v2 Announce Type: replace Abstract: Federated Self-Supervised Learning (FSSL) integrates the privacy advantages of distributed training with the capability of self-supervised learning to leverage unlabeled data, showing…</description>
</item>
<item>
<title>Post-Quantum Secure Aggregation via Code-Based Homomorphic Encryption</title><link>https://arxiv.org/abs/2601.13031</link>
<guid isPermaLink="false">veille techno-1792130487228143</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2601.13031v2 Announce Type: replace Abstract: Secure aggregation lets a server learn the sum of inputs from multiple users without learning individual contributions. Recent single-server protocols achieve this with key- and…</description>
</item>
<item>
<title>Password Strength Detection via Machine Learning: Analysis, Modeling, and Evaluation</title><link>https://arxiv.org/abs/2505.16439</link>
<guid isPermaLink="false">veille techno-1792130487228144</guid>
<pubDate>Fri, 16 Oct 2026 04:00:00 +0000</pubDate>
<description>arXiv:2505.16439v3 Announce Type: replace Abstract: As network security issues continue gaining prominence, password security has become crucial in safeguarding personal information and network systems. This study first introduces various…</description>
</item>
</channel></rss>