Black Forest Labs has introduced FLUX 3, a new foundation model designed to generate images and video with synchronized audio, while also serving as a base for robotics systems. The company is initially limiting access to selected users and partners through an Early Access programme. FLUX 3 Video can create clips up to 20 seconds …
Anthropic has expanded Claude’s voice mode to include its more capable Opus and Sonnet models, alongside the faster Haiku model. The update is designed for longer, more complex spoken conversations, such as preparing a client pitch, testing an idea or reviewing a decision. Users can select a model during a voice conversation or let Claude …
A federal judge in San Francisco has given final approval to Anthropic’s $1.5 billion settlement with authors and publishers whose books were obtained from pirate websites. The deal resolves a major copyright class action over material connected to the training of Anthropic’s Claude AI models. The settlement is the largest known recovery in a US …
Google’s expansion of AI-powered search is raising fresh concerns that the company could weaken the open web by keeping users inside its own products. Kate Conger reports for The New York Times that publishers, retailers and other website operators are seeing search traffic fall as Google provides more answers directly through AI. Google’s AI Mode …
Kimi has introduced Kimi Work, a desktop application for macOS and Windows that combines AI chat with direct access to local files, browser automation and scheduled tasks. Kimi announces the product in an official blog post, positioning it as a tool for longer workflows that end in completed files, reports or code rather than a …
Access to AI advice can make people less accurate while dramatically increasing their confidence, according to a study by researchers from universities in France and Italy. Participants were far less likely to acknowledge uncertainty after receiving AI input, even when the model gave unreliable answers. Ana Maria Constantin reports for The Next Web that the …
Qwen has introduced Qwen-Image-3.0, a new image-generation model designed for information-heavy visuals, small text and realistic details. QwenTeam announces the model in an official Qwen blog post, describing its central goal as making generated images more useful in practical work. The company says Qwen-Image-3.0 accepts instructions of up to 4,500 tokens. That capacity is intended …
Microsoft is making a multibillion-dollar commitment to expand Mistral’s AI computing infrastructure in Europe, as the two companies broaden their partnership for enterprises and regulated industries. The agreement gives Microsoft additional European capacity for its cloud and AI services. It also gives customers more ways to use Mistral’s models while keeping workloads, data and operations …
Google is expanding its Gemini model family with two lower-cost general-purpose models and a cybersecurity-focused system that is initially limited to governments and trusted partners. The releases underline Google’s effort to compete on the price, speed and operating efficiency of AI tools, particularly for companies building automated workflows. The most capable of the new public …
Powerful AI models that companies can download, run and adapt are becoming a serious competitive force against the closed systems offered by OpenAI and Anthropic. A new wave of open-weight models from China, including Moonshot AI’s Kimi K3 and Alibaba’s Qwen 3.8, has sharpened a debate in Washington and Silicon Valley over whether open models …