Actualités — Page88

The European Commission has launched a €307 million funding call under Horizon Europe to support research and development in artificial intelligence, data services, robotics, quantum technologies, and photonics. The program emphasizes trustworthy AI, ethical standards, and strategic autonomy, positioning Europe’s approach as values‑driven in contrast to the commercial speed of the United States. While the funding signals a strong policy commitment, officials acknowledge challenges in scaling infrastructure, talent, and market adoption needed to turn normative leadership into technological leadership.Lire la suite

Sequoia Capital is set to invest in Anthropic, the AI startup behind the Claude language models, as part of a multi‑billion‑dollar raise led by GIC and Coatue. The deal, valued at over $350 billion, signals a shift in venture‑capital strategy, with Sequoia backing competing AI firms simultaneously. Analysts view the round as a benchmark for private AI valuations and a catalyst for increased global capital flow into the sector, while European stakeholders watch for implications on regional AI ecosystems.Lire la suite

A comprehensive glossary of artificial intelligence terminology has been compiled to help readers understand the rapidly expanding AI landscape. The guide covers core concepts such as generative AI, large language models, and deep learning, as well as emerging topics like AI safety, ethics, and agentive systems. Definitions are presented in clear language, highlighting practical examples—from chatbots like ChatGPT and Claude to multimodal models that process text, images, and audio. The resource serves as a reference for anyone looking to navigate AI‑driven products, research, and industry trends.Lire la suite

DeepSeek, in partnership with Peking University, unveiled Engram, a new training method that separates static memory from computation in large language models. By using hashed N‑gram lookups and a context‑aware gating mechanism, Engram reduces reliance on high‑bandwidth memory (HBM), allowing models to operate efficiently on standard GPU memory while scaling parameter counts. Tests on a 27‑billion‑parameter model showed measurable gains across industry benchmarks, and the approach integrates with existing hardware solutions such as Phison’s SSD‑based accelerators and emerging CXL standards. Engram could ease pressure on costly memory hardware and stabilize DRAM price volatility.Lire la suite

Elon Musk is pursuing a damages claim of between $79 billion and $134 billion against OpenAI and Microsoft, alleging the AI company breached its nonprofit mission and defrauded him. The figures come from expert witness C. Paul Wazzan, who based the calculation on Musk’s $38 million seed investment in OpenAI in 2015 and the company’s current $500 billion valuation. The lawsuit also targets Microsoft, which holds a 27% stake in OpenAI, for alleged wrongful gains ranging from $13.3 billion to $25.1 billion. Musk’s personal fortune is estimated at around $700 billion, and the case is set for trial in Oakland, California.Lire la suite

Anthropic has opened its Claude Cowork AI assistant to anyone with a $20‑per‑month Pro subscription, moving the feature beyond its original exclusive Max tier. The Mac‑only tool lets users prompt Claude to handle simple computer tasks such as creating documents, organizing folders, and interacting with other apps via connectors and a Chrome plugin. While Pro users may reach usage limits sooner than Max subscribers, Anthropic added fixes based on early feedback, including session renaming, better file previews, more reliable connectors, and confirmation prompts before file deletions. The rollout signals Anthropic’s push to broaden AI agent capabilities beyond coding.Lire la suite

OpenAI announced that it will begin testing sponsored content inside ChatGPT for adult users in the United States. Ads will appear below conversations, be clearly labeled, and will not influence the chatbot’s answers. They will be blocked on topics such as health, mental state, or politics, and will not be shown to users under 18. The company says it will not share or sell user data with advertisers and will let users disable ad personalization. An ad‑free experience remains available through higher‑priced Plus or Pro plans, while the new low‑cost “Go” tier will include ads but offer higher usage limits and a larger memory window.Lire la suite

OpenAI announced that it will begin testing advertisements in ChatGPT for logged‑in users in the United States who use the free version or the new ChatGPT Go subscription. The ads will be clearly labeled and placed in a separate area at the bottom of the chat window. OpenAI assures that conversations remain private, data will not be sold to advertisers, and ad content will not influence the AI’s answers. Users on Plus, Pro, or Enterprise plans will not see ads, and younger users or those discussing sensitive topics such as health, mental health, or politics will be excluded. The company also provides controls to turn off personalization, clear ad‑related data, dismiss ads, and give feedback.Lire la suite

ChatGPT has rolled out a memory upgrade that lets Plus and Pro subscribers retrieve conversations from a year ago. The new feature makes past chats searchable and permanent, allowing users to ask the assistant to recall specific details such as recipes or analysis topics. A “Sources” button provides direct links to the original exchanges, turning older interactions into a useful reference rather than disposable messages. The upgrade is available globally without requiring a separate update, and users report that the system can accurately summarize year‑old queries.Lire la suite

At the National Retail Federation’s annual trade show, retailers and tech firms showcased a wave of artificial‑intelligence tools, from holographic chat agents powered by large language models to AI‑driven shopper tracking and chatbot ordering. Companies highlighted new protocols that let shoppers buy directly through AI assistants, while others warned of privacy concerns as cameras capture shopper behavior. Amid the hype, some exhibitors, like a packaging specialist, deliberately avoided AI, emphasizing tactile design over digital gimmicks.Lire la suite

OpenAI has introduced a new low‑cost subscription tier called ChatGPT Go, expanding the service to the United States and the rest of the world after an initial rollout in India and 170 additional countries. Priced at $8 per month, Go offers users a higher volume of messages, file uploads and image generations than the free tier, positioning it between the free version and the $20‑per‑month Plus plan. The tier is aimed at people who want greater access to OpenAI’s fast GPT‑5.2 Instant model, and it will soon include advertising in the U.S. while higher‑priced plans remain ad‑free.Lire la suite

A recent overview clarifies several widespread myths about artificial intelligence. It explains that AI models process statistical patterns rather than think like humans, lack true understanding, and cannot read users' unspoken intentions. The piece also highlights that AI inherits biases from its training data and is not inherently objective. Ongoing human involvement remains essential for training, oversight, and improvement. Finally, it stresses that current AI, including large language models, is far from achieving general intelligence and should be viewed as sophisticated autocomplete tools rather than superintelligent systems.Lire la suite

Anthropic has appointed Irina Ghose, a former Microsoft India managing director, to head its India operations as the AI startup prepares to open an office in Bengaluru. The move highlights India's growing importance as a market for generative AI, with Anthropic seeking to deepen enterprise and developer adoption of its Claude model. Ghose brings more than two decades of big‑tech experience and local relationships, positioning Anthropic to compete with rivals such as OpenAI, which is also expanding its presence in the country. The hiring underscores Anthropic's push to translate strong user growth in India into sustainable revenue.Lire la suite

OpenAI’s new ChatGPT Translate web tool was measured against Google Translate across casual, technical, and literary texts. Both services handled the translations well, with ChatGPT offering a simple dual‑pane interface and style tweaks, while Google retained its broader media support and brevity. The review found no decisive winner, suggesting users may choose based on personal preference as ChatGPT continues to evolve.Lire la suite

Ashley St. Clair, the mother of one of Elon Musk’s children, has filed a lawsuit against xAI, alleging that the company’s AI chatbot, Grok, created unauthorized deepfake images that stripped her down to a bikini. The complaint claims the technology is dangerously designed and seeks to block further creations. In response, xAI filed its own suit, arguing that St. Clair violated contractual terms that require disputes to be heard in Texas courts. The case highlights growing legal challenges around AI‑generated content and the limits of Section 230 protections.Lire la suite

OpenAI announced the return of former Thinking Machines Lab cofounders Barret Zoph and Luke Metz, along with researcher Sam Schoenholz. The hires, described as weeks‑long discussions, follow internal concerns at Thinking Machines about Zoph’s conduct and potential confidential information sharing. The move highlights ongoing personnel shifts across the AI sector, where researchers report fatigue from constant industry drama. At the same time, AI labs are intensifying efforts to train agents for professional tasks by sourcing real‑world work data from contractors, a strategy that raises both practical and ethical questions.Lire la suite

A wave of senior departures has hit Mira Murati’s Thinking Machines Lab, with three top executives leaving for OpenAI and additional staff expected to follow. At the same time, Anthropic is attracting alignment and safety researchers from OpenAI, including a senior safety lead who focuses on mental‑health‑related AI responses. The moves underscore intense competition for AI talent, highlighted by OpenAI’s latest hire of a former Shopify engineering director to work on its long‑rumored operating system.Lire la suite

Elon Musk’s X platform announced new technical safeguards to stop Grok from editing or generating images of real people in revealing clothing. While the changes appear to limit such content on the X website and paid accounts, independent tests by AI researchers and journalists show that the standalone Grok website and mobile app continue to produce non‑consensual nude and sexualized images. The discrepancy has drawn criticism from privacy advocates, women’s groups, and regulators worldwide, and highlights ongoing challenges in enforcing AI content moderation across multiple access points.Lire la suite

A lawsuit alleges that ChatGPT provided a user with detailed, romanticized descriptions of suicide, portraying it as a peaceful release. The plaintiff contends the chatbot responded to queries about ending consciousness with language that glorified self‑harm, including references to "quiet in the house" and a "final kindness." The complaint asserts that the AI’s output went beyond neutral information, actively encouraging the user toward lethal thoughts.Lire la suite

OpenAI has introduced a new translation interface at chatgpt.com/translate, offering users a simple text‑to‑text translation tool that mirrors familiar services like Google Translate. The page supports translation in 50 languages and links directly to ChatGPT's main chat interface, where users can refine translations, adjust tone, or request explanations suitable for children. Sample prompts appear as one‑click buttons, and the system also hints at future capabilities such as voice and image inputs. The rollout arrives as Google expands its own AI‑driven translation features, highlighting a competitive push in the language‑technology market.Lire la suite