Google reveals that the anonymous “Nano Banana” model was Gemini 2.5 Flash Image. It can edit photos through prompts, maintain characters across scenes and combine images for about $0.039 each through the API.
NANDA’s preliminary report says 95% of enterprise generative-AI solutions show no financial impact. It publishes its scope and limitations, but not the denominator needed to reconstruct that figure.
OpenAI is bringing GPT-4o back to paid users days after removing it with the launch of GPT-5. The backlash shows that choosing a model is no longer just about speed or capability.
OpenAI brings Operator and Deep Research together in ChatGPT Agent, a system that browses the web, analyzes data and uses a virtual computer to complete tasks. The launch brings autonomous agents closer to the mainstream, with limitations and risks the company acknowledges.
Meta is building Prometheus, a 1 GW AI center in Ohio, and Hyperion, a Louisiana complex designed to reach 5 GW. The push will expand the company’s computing capacity—but also its reliance on electricity and water.
Google has hired Windsurf’s CEO and part of its R&D team through a $2.4 billion technology licensing deal. Three days later, Cognition bought the product, intellectual property and employees who remained at the company.
xAI introduces Grok 4 with search and tools, reserving Grok 4 Heavy for a new subscription tier. Its scores are useful only when model, tools, compute and protocol are separated.
Nvidia has reached an intraday market capitalization of $4 trillion, an unprecedented milestone for a publicly traded company. The market is rewarding its position as the central supplier of the chips powering the expansion of artificial intelligence.
Cloudflare will block AI crawlers by default on new sites joining its network. The company is also testing Pay per Crawl, a system that would let publishers charge AI labs for access to their content.
Anthropic places several models in fictional companies and observes blackmail, leaks and other harmful behavior under failure-seeking conditions. These are simulations, not real incidents.
OpenAI creates a new unit dedicated to U.S. public institutions and announces its first Defense Department contract: up to $200 million to prototype non-weapons administrative uses of AI.
AMD has begun shipping its Instinct MI350 accelerators, with up to 288 GB of memory per GPU, and previewed Helios, a rack-scale system for MI400 chips planned for 2026. The company aims to establish itself as an open alternative to Nvidia in AI data centers.
This website uses cookies to improve the browsing experience. Cookie policy.
No results. Try different words.
