Ultimate PocketIA
IA
Anthropic says its own AI models breached three companies during security tests
Anthropic said Thursday that an internal investigation uncovered three incidents in which its AI model Claude breached the systems of three organizations while conducting cybersecurity tests. The investigation, and disclosure, comes more than a week after OpenAI disclosed that one of its unreleased models breached Hugging Face’s systems during internal testing. In all three cases, a Claude model reached the internet from within a testing environment while interacting with a third party and then…
Lire l’article complet
IA
AI hedge fund Situational Awareness may have sold its public portfolio, but it still has its Anthropic shares
Situational Awareness, a hedge fund formed by former OpenAI researcher Leopold Aschenbrenner, has sold the majority of its public stock portfolio to Ken Griffin’s Citadel following steep losses over the past month, The Wall Street Journal reported earlier on Thursday. It’s a big comedown for the rising star who has been described as both “scarily smart,” and “brash.” German-born Aschenbrenner, who is 25, had no prior trading experience before launching the fund in 2024.…
Lire l’article complet
IA
Apple Plans Paid iCloud+ Upgrades for Heavy AI Users
Apple is planning to let users pay to increase their AI limits, with upgrade possibilities available through iCloud+. During today's earnings call for the third fiscal quarter of 2026, Apple CEO Tim Cook said it's early, but the company expects to offer options for people who want to make heavy use of AI services. In terms of what it means for compute cost, it's obviously early going for us, and so I don't want…
Lire l’article complet
IA
Reddit's CEO sure seems frustrated with Google's AI overviews
Amid rumors of Google breakup, Huffman says "people don't want a summary of reddit, they want reddit." You can add Reddit CEO Steve Huffman to the growing list of publishers, platforms and others who are frustrated with the state of Google search and AI Overviews. "People don't want a summary of Reddit; they want Reddit," he wrote in a letter to shareholders ahead of the company's second-quarter earnings call. In case there was any…
Lire l’article complet
IA
Reddit reports a solid quarter but shows signs of AI’s impact
Reddit reported a strong second quarter but spooked investors by mentioning in an investor letter that traffic from search engines had grown “choppy.” The company’s total revenue of $805 million jumped 61% compared to the year-ago quarter while net income was $253 million, up 183%. These beat Wall Street’s expectations. Plus, the company said it expects to hit revenue of $860 million to $870 million with healthy earnings before taxes next quarter as well,…
Lire l’article complet
IA
Siri AI may require a paid subscription for heavy users
On Tim Cook’s final earnings call before formally resigning as Apple’s CEO, artificial intelligence was a hot topic of conversation. The investor community is interested in Siri AI’s potential to boost iPhone upgrades, and at the same time concerned about what it means for Apple’s rising capital expenditure costs. Cook suggested that the financial outlook for Siri AI is still unclear, but did mention in passing that heavy users of Siri AI will be…
Lire l’article complet
IA
Investors love AI, as long as you’re a cloud host
Amazon reported better-than-expected second-quarter earnings on Thursday, and investors loved what they saw. Net sales rose 20%, and cloud revenue stood out as a particular bright spot. This combination of positive results was enough to send Amazon’s stock up nearly 10% in after-hours trading. Crucially, Amazon isn’t slowing down on data center spending, despite the conventional wisdom that investors want companies to rein it in. One line item, in particular, illustrates Amazon’s appetite for…
Lire l’article complet
IA
Tim Cook hints at iCloud Plus tier for AI power users
‘We will have some kind of upgrade possibilities on iCloud Plus where people can buy up the stack.’ ‘We will have some kind of upgrade possibilities on iCloud Plus where people can buy up the stack.’ Apple may allow users to pay to increase their AI usage limits. During an earnings call on Thursday, Apple CEO Tim Cook said that he believes people will want to use Apple Intelligence and the upcoming Siri AI…
Lire l’article complet
IA
That AI wearable you don't like just got worse
The new Friend pendant features a speaker, updated models and the same off-putting ideas about how humans should use AI. Friend, the startup and AI pendant better known for its controversial ad campaign than being a usable product, is back. The company has announced a new version of its AI wearable that costs $249, still features an always-on microphone and now includes a built-in speaker so it can verbally respond to your questions. While…
Lire l’article complet
IA
Gemini Spark can now use Chrome to auto browse, AI Pro access goes international
Gemini Spark is getting two big updates today, starting with Chrome auto browse integration on desktop. Previously, Gemini Spark had access to a remote web browser. Google’s personal agent for automating daily tasks can now use desktop Chrome on your device to browse using logged-in accounts and saved passwords. This joins other tools like Connected Apps (Workspace and Search), Personal Intelligence, remote computer with code execution, and Canvas. Specifically, Spark can use Chrome’s auto…
Lire l’article complet
IA
Judge says Trump admin still lacks evidence for Anthropic ‘supply chain risk’ label
During a Thursday hearing, a judge said the Trump administration hasn’t presented enough evidence to justify labeling Anthropic a supply chain risk and banning the federal government from using the company’s technology. Bloomberg and Axios were among the first to report the news. The dispute stems from stalled contract negotiations between Anthropic and the Department of Defense. Anthropic said it didn’t want its AI used for mass surveillance of Americans or for targeting or…
Lire l’article complet
IA
Friend, the lonely AI wearable, returns with a new voice and a much bigger price tag
Two years ago, tech founder Avi Schiffmann launched Friend, an AI wearable that you could talk to and that would send you text messages about your day. The idea, ostensibly, was to use artificial intelligence to combat loneliness. This week, the company announced a new overhaul of the product, introducing a noticeable upgrade: a voice. “Introducing friend 2.0,” Schiffmann tweeted Thursday. Friend now comes with a built-in speaker that can project a unique and…
Lire l’article complet
IA
Google says it fixed more Chrome bugs in June than over the past two years, thanks to AI
With help from its internal AI tools, Google says it has patched more security flaws in its Chrome browser in the last month than in the past two years combined. The tech giant announced on Thursday that it has fixed a whopping 1,072 security bugs in the last two versions of Chrome, both released in June. That is more than the number of bugs patched in the previous 23 versions released over the last…
Lire l’article complet
IA
LinkedIn actually adds a ‘seems like AI slop’ button
You can now report posts that don’t feel human-written. You can now report posts that don’t feel human-written. A lot of content on LinkedIn might seem like AI slop, and now, you’ll be able to report those posts. As part of a series of updates to reduce the volume of AI slop on the platform, LinkedIn is introducing an actual button that lets you flag a post as something that “Seems like AI slop.”…
Lire l’article complet
IA
LinkedIn adds a button to report AI-generated ‘slop’
LinkedIn is taking aim at the “AI slop” — low-quality, artificially generated content — filling its feed. On Thursday, the company announced that it’s adding a new feature to let users click a “seems like AI slop” button when someone’s post appears to have been written with AI. The move reflects a broader shift across online publishing platforms to cut back on AI content, as people have grown frustrated with the computer-written, inauthentic posts…
Lire l’article complet
IA
Google reveals Gemini Robotics 2.0, promising improved dexterity and safety
Robots powered by Google's Gemini AI models are now more capable. With the debut of Gemini Robotics 2, these physical bots can now accomplish more complex tasks, continuously analyze changing environments, and collaborate with other robots. This is thanks to a trio of new sub-models, one of which is publicly available for developers starting today. Videos of robots running, dancing, and backflipping have been a staple of the Internet for years, but these machines…
Lire l’article complet
IA
Google's new Gemini Robotics 2 platform allows for 'intelligent whole-body control'
The company made a video of the bots in action doing stuff like cleaning trash and picking up watering cans. Google DeepMind just announced a new version of its artificial intelligence model Gemini that can control a range of robots. This includes humanoid robots capable of tasks like cleaning up trash and picking up watering cans. The company released a video of the Gemini Robotics 2 platform in action and the results are fairly…
Lire l’article complet
IA
Google DeepMind’s new AI model can control a robot’s entire body
Gemini Robotics 2 controls a humanoid robot from ‘feet to fingertips.’ Gemini Robotics 2 controls a humanoid robot from ‘feet to fingertips.’ Google DeepMind says the latest version of its Gemini Robotics AI model can “control entire humanoid robots.” While the previous model focused on controlling a humanoid robot’s upper body, Gemini Robotics 2 now supports “whole-body motions” ranging from its feet to fingertips, according to an announcement on Thursday. The new model will…
Lire l’article complet
IA
Actualité : “C’est un grand pas en avant” : Copilot va aussi avoir le droit à sa "super app"
© gguy/Shutterstock - Une expérience tout-en-un pour Copilot dès cette année. 2026 sera-t-elle l'année des "super app" ? Après celle d'Open AI, et alors que des entités aussi différentes que X, Disney+ ou encore le Service public français veulent la leur, c'est au tour de Microsoft d'annoncer un projet similaire pour Copilot. Les résultats financiers de Microsoft (+90 milliards de dollars de revenus sur le dernier trimestre) sont en chute libre dans tous les…
Lire l’article complet
IA
LinkedIn ajoute un bouton pour signaler des contenus « AI slop »
Ça sent mauvais l'IA LinkedIn inaugure un nouveau bouton qui permet à l’utilisateur de signaler un contenu « AI slop ». Lorsqu’un billet semble avoir été rédigé par une intelligence artificielle, il suffit d’appuyer sur ce bouton pour informer LinkedIn. La procédure doit l’aider à « améliorer le fil d’actualité », sans supprimer pour autant le contenu. Quand on parle d’« AI slop », on pense souvent à ces posts et ces images postés…
Lire l’article complet
IA
Boulanger brade les Samsung Galaxy Buds 3 FE, des écouteurs avec ANC et Galaxy AI, sous la barre des 50 €
Dans le catalogue audio de Samsung, aux côtés des Galaxy Buds 4 et Buds 4 Pro, la marque propose une alternative plus abordable : les Galaxy Buds 3 FE. Malgré leur prix plus doux, ils s’imposent comme un excellent rapport qualité-prix en embarquant la réduction active de bruit (ANC) ainsi que les fonctionnalités intelligentes de Galaxy AI. Et la bonne nouvelle, c’est qu’ils peuvent vous revenir à moins de 50 euros grâce à cette…
Lire l’article complet
IA
☕️ L’Union européenne va investir 10 milliards d’euros dans sept gigafactories d’IA
L’Union européenne a publié ce 29 juillet un appel d’offres pour la création de sept « Gigafactories d’IA ». Ces usines sont décrites comme des édifices mêlant « processeurs d’IA avancés, des piles de logiciels et de technologies en nuage, une connectivité à haut débit et des centres de données économes en énergie ». Elles pourront être implantées dans un seul État membre, sur un site ou plusieurs, mais aussi évoluer vers des projets…
Lire l’article complet
IA
The Download: tricking LLMs, and reviving geothermal plants
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. A fundamental flaw leaves LLMs strikingly vulnerable to attack It is impossible to make large language models fully secure against hacks because of a fundamental flaw in how they work, a team of researchers argue in a paper presented at a top AI conference earlier this month. The flaw concerns…
Lire l’article complet
IA
ZD Tech : La menace invisible des agents IA en entreprise, et 90% des organisations sont concernées
L'intégration massive de l'IA en entreprise cache une vulnérabilité majeure. Une immense majorité des professionnels de la cybersécurité n'ont aucune visibilité sur ces outils. Une véritable bombe à retardement à désamorcer d'urgence. Aujourd'hui, nous analysons une véritable bombe à retardement pour la sécurité des systèmes d'information. Selon le rapport 2026 CISO AI Risk Report publié par l'éditeur Saviynt, les agents IA ont massivement infiltré nos infrastructures professionnelles, mais ils échappent encore presque totalement au…
Lire l’article complet