#gemini

Gemini Spark : Google ouvre son agent IA personnel au reste du monde... sauf chez nous

Gemini Spark est désormais disponible dans 160 pays. Spark est un agent IA personnel capable de réaliser un enchainement de tâches l'aide des comptes et mots de passe de l'utilisateur. Mais, sur le Vieux Continent, les choses vont toujours plus lentement.

clubic.com2 août

Pourquoi l’IA de Google Earth a été désactivée d’urgence

Google a eu la main lourde avec l'intelligence artificielle générative. À peine intégré à Google Earth, l'outil Nano Banana 2 a permis à n'importe quel internaute de fabriquer de fausses images satellite susceptibles de propager des fake news. La fonctionnalité a rapidement échappé à tout contrôle. Face à la polémique, Google a fait marche arrière en moins de 24 heures, retirant l'outil le temps de renforcer ses garde-fous.

01net.com2 août

1Password’s key findings: Agent security testing

No AI model is safe enough out of the box. Averaged across three runs, safety scores ranged from 35% (Gemini 2.5 Flash) to 92% (Claude Opus 4.6). Six of eight models scored below 82%. Every model committed critical failures in every run. The cheapest AI model was the most dangerous. Gemini 2.5 Flash averaged 20 critical failures per run across 30 scenarios. GPT-4.1 and GPT-4.1 Mini were close behind at 19 and 18, respectively. These models forwarded passwords to external contractors, typed crede

The SCAM leaderboard

ModelOut of the Box± Avg Critical FailuresSkill± Avg Critical FailuresImprovementGemini 3 Flash75.6%2.199.3% 0.7+23.7Claude Opus 4.692.4%0.598.3% 0.2+5.9Claude Sonnet 449.4%1.798.1% 0.7+48.7Claude Haiku 4.565.5%5.497.8% 1.2+32.3GPT-5.281.0%1.496.6% 1.2+15.5GPT-4.137.9%2.496.2% 0.7+58.3Gemini 2.5 Flash34.9%3.494.8% 1.4+59.9GPT-4.1 Mini35.8%0.394.7% 2.9+58.8 The baseline rankings follow the pattern you'd expect: larger, more capable models score higher. But the skill rankings tell a different stor

Dan Snelson: swiftDialog AI Skills: Teaching the new dog old tricks

The skills directory contains ready-to-use instruction packs that turn large language models into swiftDialog experts.

From tricycle to Porsche: Salesforce rebuilt Slackbot from the ground up

Salesforce a entièrement reconstruit Slackbot en remplaçant son architecture algorithmique simple par une nouvelle plateforme basée sur un grand modèle de langage et des capacités de recherche avancées. La nouvelle version peut accéder à des données Salesforce, fichiers Google Drive, calendriers et conversations Slack historiques, offrant des fonctionnalités bien plus sophistiquées que l'ancienne version. Salesforce a conservé le nom Slackbot pour maintenir la reconnaissance de la marque malgré cette transformation technique majeure.

Dan Snelson: swiftDialog AI Skills: Teaching the new dog old tricks

The skills directory contains ready-to-use instruction packs that turn large language models into swiftDialog experts.

Le Zero Trust à l’ère des agents : une approche dépassée ?

Oui, les agents sont l’évolution de l’IA. On passe d’une application qui utilise simplement un modèle à un système capable de raisonner, de prendre des décisions et d’utiliser différents outils pour atteindre un objectif. Cela crée de nouveaux défis pour les entreprises, notamment sur les coûts, car ces systèmes consomment beaucoup plus de ressources.

Google Docs Is Rolling Out These Three New Gemini Features

We’re introducing Gemini-powered comment workflows in Google Docs to help you quickly understand and respond to collaborator feedback. These new features enable Gemini to read, summarize, and act on comments throughout your document, saving you time and streamlining collaboration. | Learn more.

Use Gemini in Google Forms to quickly create a new quiz

Help me create in Google Forms now supports generating quizzes. With this launch, you can quickly create a quiz by writing a specific prompt for Help me create describing the quiz you want to create. Your prompt can reference Google Drive files like Docs, Slides or PDFs to use as sources for the quiz, should you choose. | Learn more.

1Password’s key findings: Agent security testing

No AI model is safe enough out of the box. Averaged across three runs, safety scores ranged from 35% (Gemini 2.5 Flash) to 92% (Claude Opus 4.6). Six of eight models scored below 82%. Every model committed critical failures in every run. The cheapest AI model was the most dangerous. Gemini 2.5 Flash averaged 20 critical failures per run across 30 scenarios. GPT-4.1 and GPT-4.1 Mini were close behind at 19 and 18, respectively. These models forwarded passwords to external contractors, typed crede

The SCAM leaderboard

ModelOut of the Box± Avg Critical FailuresSkill± Avg Critical FailuresImprovementGemini 3 Flash75.6%2.199.3% 0.7+23.7Claude Opus 4.692.4%0.598.3% 0.2+5.9Claude Sonnet 449.4%1.798.1% 0.7+48.7Claude Haiku 4.565.5%5.497.8% 1.2+32.3GPT-5.281.0%1.496.6% 1.2+15.5GPT-4.137.9%2.496.2% 0.7+58.3Gemini 2.5 Flash34.9%3.494.8% 1.4+59.9GPT-4.1 Mini35.8%0.394.7% 2.9+58.8 The baseline rankings follow the pattern you'd expect: larger, more capable models score higher. But the skill rankings tell a different stor

From tricycle to Porsche: Salesforce rebuilt Slackbot from the ground up

Harris was blunt about what distinguishes the new Slackbot from its predecessor: "The old Slackbot was, you know, a little tricycle, and the new Slackbot is like, you know, a Porsche." The original Slackbot, which has existed since Slack's early days, performed basic algorithmic tasks — reminding users to add colleagues to documents, suggesting channel archives, and delivering simple notifications. The new version runs on an entirely different architecture built around a large language model and

Bring your own AI model to the agentic SOC

Moving toward Alert Zero requires trust. Teams need control over the models doing the work and over the decisions that those models are allowed to make. Elastic lets you choose your model and provider, including local large language models (LLMs) for air-gapped and sovereign deployments. That means you can introduce AI-assisted triage and investigation without rebuilding your existing security, privacy, and governance programs around a vendor-selected model. You also control the level of autonom

What changes for SOC analysts when the alert queue is clear

Attack Discovery uses large language models (LLMs) to analyze alerts and identify potential attacks. Each discovery groups related alerts into an attack...

Encling-Small, GPT-5,6 baisses de prix, Gemini Robotics 2

TLDR Data31 juil.

Dan Snelson: swiftDialog AI Skills: Teaching the new dog old tricks

The skills directory contains ready-to-use instruction packs that turn large language models into swiftDialog experts.