Realtime AI News
OpenAI limits the rollout of GPT-5.6 following a government request, stating that such government access processes should not become the long-term default.
Google Research Blog publishes an article on accelerating Gemini Nano models on Pixel devices using frozen multi-token prediction.
OpenAI has hired Uber's India head to lead its operations in India, its largest market outside the United States, signaling deeper local investment.
An AI agent reportedly survived 6,000 hacking attempts in security testing, as reported by Decrypt.
Qualcomm, OpenAI, and IBM are jointly targeting AI infrastructure efficiency improvements, as reported by TechTarget.
OpenAI has shared plans for Jalapeño, a custom AI inference chip developed with Broadcom, joining Google, Apple, and SpaceX in a growing movement to reduce dependence on Nvidia's dominant GPU lineup.
Reuters and Axios report that OpenAI has deferred the public deployment of GPT-5.6 after the U.S. government requested early access to the frontier AI model.
OpenAI has released GPT-5.6, its latest AI model, while the US regulatory battle over AI framework legislation enters a critical phase.
OpenAI's decision to postpone its IPO has sent SoftBank's stock crashing 13%, putting the Japanese investment giant's $65 billion exposure under severe liquidity pressure.
The US government has told OpenAI to restrict access to its most powerful AI model, as reported by Computerworld.
OpenAI has decided to postpone its Initial Public Offering (IPO) plans, according to a report from blockchain.news.
Chinese physical AI company DeepWisdom closed two consecutive funding rounds totaling hundreds of millions of yuan within two months, pushing forward its full-stack self-developed approach to physical world AI foundation models.
The Trump administration has restricted OpenAI from releasing its latest GPT-5.6 model, reportedly due to concerns over its capabilities being too advanced.
A Chinese AI model is reportedly becoming a new competitive threat to Anthropic as the company prepares for its planned IPO, according to Yahoo Finance UK.
OpenAI has revealed plans for Jalapeño, its custom AI inference chip developed in partnership with Broadcom, joining a growing list of tech giants building their own silicon.
SoftBank shares tumble as market uncertainty grows over the timeline of OpenAI's IPO, with the Japanese conglomerate being a major investor.
A new Nature study reveals that people can develop real emotional connections while chatting with Claude, with the AI being described as the 'new generation electronic husband.'
A 3D generation company backed by hundreds of millions in funding achieves a breakthrough — producing millions of faces in 4 seconds with 10M+ precision and 12K HD textures.
SignalFire data reveals that despite AI dominating layoff narratives, engineers are actually making up a larger share of new hires, becoming one of the most resilient job groups.
Top AI researchers Jonas Adler and Alexander Pritzel have left Google to join Anthropic, following the departures of Noam Shazeer and John Jumper.
3D vision perception company Orbbec demonstrated its latest AI-driven vision systems at the Automate 2026 trade show.
HIVE Digital Technologies has signed a Letter of Intent to deploy up to 10,000 NVIDIA GPUs at a data center in Boden, Sweden.
According to ibtimes.sg, AI-related regulatory fines have already reached $3.5 billion globally, with the most significant legal challenges yet to come.
According to qz.com, enterprise customers are reducing their spending on AI services from OpenAI and Anthropic, reflecting a reassessment of AI return on investment.
Google announced on its official blog that the Interactions API will become the primary interface for Gemini models and agents.
Anthropic formally accused Alibaba of harvesting approximately 28.8 million Claude chat conversations, alleging AI model data theft.
Cloud infrastructure company Nscale announced Alfred, an AI agent built for modern engineering teams.
Bloomberg reports that OpenAI pitched its advertising business at the Cannes Lions International Festival of Creativity, signaling a major push into ad-driven revenue.
According to Security Insider (安全内参), AI-powered attack automation has driven a nearly 15-fold surge in phishing attacks in 2026.
OpenAI previewed GPT-5.6 Sol on June 26, describing it as a next-generation model with enhanced capabilities in coding, science, and cybersecurity, paired with its most advanced safety stack.
Hyundai has officially unveiled the new-generation Avante/Elantra, featuring integrated AI technology and a bold angular design language.
The AI prediction ecosystem platform 'manadia' has officially launched worldwide, aiming to bridge the gap between AI technology potential and realized value.
China has officially published a series of 7 national standards titled 'Artificial Intelligence — AI Agent Interconnection,' establishing a standardization framework for interoperability between AI agents.
Anthropic's internal salary data reveals technical staff can make over $1 million annually, according to a Business Insider report.
OpenAI's IPO has been delayed, while SpaceX is considering entering the mobile network business.
Qwen team officially published Qwen3-ForcedAligner-0.6B-hf on Hugging Face, a forced alignment model supporting Chinese, English, Cantonese, and French.
Qwen team launched Qwen3-ASR-0.6B-hf on Hugging Face, an automatic speech recognition model supporting Chinese, English, and Cantonese.
The Qwen team has officially released the Qwen3-ASR-1.7B model on Hugging Face, focusing on automatic speech recognition with support for Chinese, English, and Cantonese.
Assort Health has raised $120 million to build a voice AI agent platform targeting the healthcare industry.
Nvidia CEO Jensen Huang acknowledged that revenue from China has dropped to zero and unveiled a $20 billion recovery plan.