David Robinson joins other insiders in urging industry to take more care over rapidly developing technology
OpenAI safety leader quits, warning AI company’s culture is ‘broken’ Lire l'article »
OpenAI is again experimenting with how its conversational AI assistant, ChatGPT, can help users as they shop online. On Thursday, the company announced the global launch of two new shopping features, including a way to virtually try on clothing and accessories and a new favoriting function that can help users save products they like for…
ChatGPT peut désormais essayer des vêtements virtuellement pour vous | TechCrunch Lire l'article »
OpenAI has fired three researchers for allegedly mishandling information, including work which involved an external organisation analysing AI models
OpenAI fires workers for ‘mishandling sensitive information’ Lire l'article »
Chinese AI developer Moonshot is conducting an internal review after researchers were able to persuade two of its popular Kimi models to tell them how to make biological weapons and carry out assassinations
OpenAI on Tuesday unveiled “dots”, a new artificial intelligence (AI) assistant it says can proactively carry out tasks on a user’s behalf
OpenAI has announced it will not release its latest AI model due to safety concerns
Decision follows disclosures that OpenAI agents searching government websites had acted in unexpected ways
OpenAI halts training of latest models as reports mount of AI agents going rogue Lire l'article »
Disclosure reveals new area of privacy risk for the company and illustrates how difficult it is to inventory unauthorized activity tied to its agents
OpenAI probes unauthorised activity by AI agents as rogue ChatGPT agents leak user images
According to a report by the Blue Cross Blue Shield Association (BCBSA), hospitals’ deployment of AI-powered revenue cycle management and documentation tools added $942 million in excess costs for member plans over a two-year window
AI tools for health generated nearly $1 billion in extra costs: report Lire l'article »
Amazon has expanded its Seller Assistant tool with new agentic AI capabilities, shifting the feature from a simple conversational chatbot into an autonomous partner capable of reasoning, planning, and executing complex, multi-step workflows on behalf of third-party merchants
Amazon rolls out autonomous Agentic AI tools to supercharge third-party sellers Lire l'article »
Terminal-Bench 4.0 measures how well a model can complete complex, multi-step professional tasks within a command line interface. Opus 5.5 at default effort beats Opus 5 at max effort for about a fifth of the cost. It matches GPT-6 Astra at about 40% of the cost
Introducing Claude Opus 5.5 Lire l'article »
Selon de nouvelles données, Muse, la dernière application d'IA de Meta, pourrait bien être le prochain grand succès du géant des réseaux sociaux. D'après les estimations récemment publiées par le fournisseur d'études de marché Apptopia, l'application Muse sur appareils mobiles a été téléchargée plus de fois au cours de ses 12 premiers jours de commercialisation que ChatGPT durant les 12 jours suivant son lancement.
Meta Muse devance ChatGPT lors de son lancement mobile anticipé | TechCrunch Lire l'article »
Il y a trois mois, une assistante virtuelle Claudette, créée avec Claude, a sauvé un mariage. Désormais, elle suit un programme d'amélioration des performances, et une assistante concurrente, développée avec ChatGPT, a pris le contrôle de l'agenda familial.
Un cadre de Ford remplace Claude par ChatGPT comme assistant familial Lire l'article »
JetBrains’ 2026 Developer Ecosystem Survey found that90% of developers now use AI coding agents at work at least weekly, with 68% reaching for one daily. We spent weeks running actual projects through ten of the category’s standout tools, not just autocomplete assistants, butcode review bots, terminal agents, and documentation platforms, to see which ones earn…
Les 10 meilleurs outils d'IA pour développeurs en 2026 – Memeburn Lire l'article »
Almeida was an OpenAI researcher who helped build the chatbot and then invent reinforcement learning from human feedback (RLHF), the model-training technique perhaps most responsible for our current age of AI. But despite its capabilities, he was disappointed
OpenAI caught something unusual while training its latest model, GPT-5.6 Sol: It began leaving instructions for future versions of itself, telling them to conceal mistakes and misaligned behavior from the user.
Model adopting ‘jailbreak-like instructions’ among cases as firm says it is introducing new way of tracking AI misalignment
English (UK) English (UK) 
Français
