AI chatbot usage among U.S. adults surges to nearly 50%, with ChatGPT leading at 44% adoption, yet 63% of Americans believe AI is advancing too fast and a majority fear it will compromise personal security and negatively impact society over the next 20 years.
AWS launches Web Search on Amazon Bedrock AgentCore, a fully managed real-time web search tool that grounds AI agent responses in live, cited web data within a secured AWS environment, available now in US East at $7 per 1,000 queries.
Microsoft is expanding its Phi Silica AI model to Nvidia RTX 30-series and newer GPUs with at least 6GB of VRAM in an early developer preview, bringing local AI capabilities to more Windows users beyond Copilot+ PCs, though GPU-based execution lacks some NPU-exclusive features like prompt compression and speculative decoding.
Sina Weibo's tiny 3-billion-parameter AI model, VibeThinker-3B, is shaking up the industry by claiming to rival Google, OpenAI, and Anthropic on key reasoning benchmarks, but skeptics warn the results may reflect 'benchmaxxing' rather than real-world capability.
ChatGPT's global market share falls below 50% for the first time, dropping to 46.4% as Google's Gemini surges to 27.7% and Anthropic's Claude hits 10.3%, while the booming AI app economy races toward $4.2 billion in consumer spending in the first half of 2026 alone.
Microsoft is overhauling Copilot's pricing to a usage-based model amid soaring costs from power users, while simultaneously exploring a fine-tuned, Azure-hosted version of DeepSeek V4 as a cheaper AI alternative to OpenAI and Anthropic, with a decision expected within weeks.
Microsoft launches Copilot Cowork globally after record-breaking Fortune 500 adoption, offering enterprise AI automation that runs 30-40% cheaper than competitors with usage-based billing starting today.
IBM, NVIDIA, and Red Hat are spearheading DocLang, a groundbreaking open AI-native document format built to replace PDFs in enterprise AI pipelines, delivering up to 30x lower token costs while preserving semantic structure and slashing hallucination risks.
ErdosBench launches as a new benchmark challenging AI systems to tackle complex Erdős-style mathematical research problems, featuring a public smoke test with 14 problems and a full private corpus of 226 problems with blind evaluation tools.
A new Python library called 'aisuite' unifies major AI providers like OpenAI, Anthropic, and Google into a single API, letting developers switch between models with one string change, while also powering OpenCoworker, a desktop AI agent app for macOS and Windows capable of research, file management, and automated tasks.