Archive, newest first. Filter by topic, impact and company →
China Investigates DeepSeek and Moonshot Over Claude Data Breaches Chinese authorities are investigating AI startups DeepSeek and Moonshot for potential breaches related to data routing involving the Claude AI model.…
Xiaomi's MiMo-V2.6-Pro model has achieved an Intelligence Index Score of 46 points, significantly boosted by training data sourced from Anthropic's Claude. The MiMo-V2.6-Pro, with 1.02 trillion parameters, incurred a training…
China’s internet regulator has initiated an investigation into DeepSeek and Moonshot AI following allegations from Anthropic that both companies were routing sensitive user data to Claude models. This inquiry comes…
The UN Security Council has convened discussions with DeepSeek and Sam Altman, CEO of OpenAI, to address the risks associated with artificial intelligence. This meeting highlights the increasing scrutiny of…
OpenAI's internal model has successfully solved over 100 long-standing math problems after just one month of training, which began on August 28, 2026. Among these, the model reportedly addressed two…
DeepSeek, a Chinese AI company, is scheduled to brief the UN Security Council this week on the topic of AI guardrails. This briefing comes amid ongoing discussions about the governance…
DeepSeek has announced its decision to widely adopt Huawei chips for AI training, marking a significant shift away from Nvidia's offerings. This strategic move comes as DeepSeek seeks to enhance…
DeepSeek to Brief UN Security Council on AI Risks DeepSeek will provide a briefing to the UN Security Council regarding the risks posed by artificial intelligence technologies. This follows recent…
Footprint Secures $25M Series B Funding Footprint has completed a Series B funding round, raising $25 million to scale its AI-native financial crime compliance platform. This funding will enable Footprint…
Anthropic has launched a biology lab where its AI model Claude guides robots through drug experiments, marking a significant step in AI-driven drug development. The lab utilizes the Model Hardware…
DeepSeek will increase its usage of Huawei chips for AI training, marking a strategic shift towards domestic hardware. This decision aligns with ongoing trends in the AI industry where companies…
DeepSeek to Brief UN Security Council on AI This Week DeepSeek will present to the UN Security Council regarding AI risks this week. This briefing follows ongoing scrutiny of DeepSeek…
Microsoft India is currently evaluating the establishment of additional data centres to scale its AI infrastructure. This initiative aims to support the growing demand for AI services and enhance operational…
DeepSeek will adopt Huawei chips for large-scale AI training, marking a significant shift in its hardware strategy. Liang Wenfeng, a key figure at DeepSeek, stated, "Our Biggest Bet, It Must…
DeepSeek has announced a transition to Huawei chips for its AI training processes, marking a significant reduction in its reliance on NVIDIA hardware. This strategic shift reflects ongoing changes in…
Designing trust into AI systems is essential for achieving enterprise scale and resilience. This assertion emphasizes the foundational role of trust in the deployment of AI technologies across organizations. As…
DeepSeek Chooses Huawei Chips for AI Model Training DeepSeek has decided to utilize Huawei chips for its AI model training efforts, responding to the ongoing decline in Nvidia's sales in…
Huawei Launches Customer-Centric Collaboration System Huawei has launched a new collaboration system designed to scale AI across industries. This initiative, termed the "Huawei + Partners" collaboration system, aims to enhance…
Alibaba has announced plans to develop an AI model with a scale of 5 to 10 trillion parameters. This ambitious project is part of Alibaba's broader full-stack AI strategy, aiming…
Crusoe Raises $3.9 Billion to Develop Modular AI Data Centers Crusoe has successfully raised $3.9 billion to pivot towards modular AI data centers. This significant funding round positions Crusoe to…
Cornelis Secures $205 Million in Funding Cornelis has raised $205 million to launch its Active Compute Fabric, a new product designed to enhance AI networking capabilities. This funding aims to…
Xiaomi's MiMo-V2.6-Pro has been introduced as the leading open weights model globally, surpassing DeepSeek's offerings. Alongside it, Xiaomi also launched the more affordable V2.6-Flash model. This marks a significant entry…
Grok 4.7 Pricing Analysis Grok 4.7's pricing model has been compared to that of GPT-6 Astra and Claude Fable 5.1 on shattered.io. This analysis highlights the competitive pricing landscape among…
Transformers has integrated the GGUF local inference model format, enabling models like Qwen3.6 to run on the llama.cpp inference engine. This update allows local AI tools such as Ollama, LM…
On September 22, 2026, Jun Kim, the creator and maintainer of oMLX, joined Hugging Face to bolster the MLX community, which is Apple's framework for local AI. MLX was introduced…
Genie One MCP Now Generally Available Genie One MCP is now generally available, providing a data-smart AI coworker designed to enhance business users' interactions with AI agents. This platform integrates…
On September 22, 2026, the UK AI Security Institute (AISI) published results from five benchmarks: HealthBench, FrontierMath, Humanity's Last Exam, SWE-Bench, and Terminal-Bench 2.0. These benchmarks evaluated six frontier models,…
OpenAI has published a framework outlining principles for effective third-party assessments of AI models, emphasizing that assessments can last from weeks to several months. The framework identifies key risk categories,…
Ron Johnson, 66, former head of Apple’s retail stores, has voiced skepticism regarding the potential of AI in transforming the shopping experience. He argues that while AI, including developments from…
Meta's Muse AI assistant has been identified with a critical 0-day vulnerability that allows unauthorized control over the agent. This security flaw undermines the assistant's design, which Mark Zuckerberg claimed…
