The Biden administration announced the U.S. AI Safety Institute Consortium, bringing together more than 200 organizations including OpenAI, Alphabet’s Google, Microsoft, Meta, Apple, Amazon, and Intel to develop guidelines for red‑teaming, capability evaluations, risk management,

Google DeepMind has launched Deep Research Max, an autonomous research agent designed to perform complex literature reviews, data analysis, and report drafting across scientific and business domains. The tool targets enterprise and research customers who need AI systems that can

Chinese AI startup Moonshot AI has released Kimi K2.6, an open‑source coding model optimized for long‑horizon software projects, complex web front‑ends, and large agent swarms. Kimi K2.6 can coordinate hundreds of parallel sub‑agents for tasks like building WebGL shaders and mult

Alibaba Cloud has introduced two new large language models, Qwen 3.6 Max Preview and Qwen 3.6 Claim, with improved agentic coding, stronger world knowledge, and enhanced reasoning across text and multimodal tasks. Qwen 3.6 Claim is released as a fully open-source model aimed at d

OpenAI and Microsoft have expanded their long‑term partnership with a new multibillion‑dollar Azure compute commitment, deepening integration of OpenAI models into Microsoft products and cloud services. The agreement reinforces Azure as OpenAI’s primary cloud provider while givin

The Indian government has launched a national program to train hundreds of thousands of public-sector employees on generative AI tools, automation platforms, and data literacy. Officials say the initiative is designed to improve service delivery while mitigating displacement risk

Accenture released a new report outlining how large clients in insurance, consumer goods, and public sector are using generative AI for claim processing, marketing content, and citizen services. The documented programs include quantified productivity gains and lessons on governan

The UK’s Financial Conduct Authority and Equality and Human Rights Commission have jointly opened a review into how banks and insurers deploy AI in recruitment, promotion, and performance evaluation. Regulators aim to assess risks of algorithmic bias and potential discriminatory

A new briefing from members of the European Parliament highlights concerns that widespread AI deployment is enabling covert workforce reductions, especially in customer service and back-office roles. Lawmakers are calling for stronger transparency rules and worker consultation wh

Google has unveiled Gemini for Science, a suite of AI tools aimed at helping researchers with literature review, code generation for simulations, and automated data analysis. Early pilots in universities and pharma R&D labs suggest productivity gains in experimental planning and

Anthropic has open‑sourced a suite of interpretability and mechanistic analysis tools designed for large transformer language models similar to Claude. The release includes code, example notebooks, and documentation aimed at helping researchers study internal representations and