Track model releases, AI agents, open-source projects, infrastructure, policy, and product updates as they publish.
Moonshot AI's Kimi product is accelerating its international expansion, choosing not to compete on price but rather on product strength and technical capability against the three dominant overseas AI companies.
HP has entered a strategic partnership with OpenAI to accelerate the development and deployment of the AI agent platform Frontier.
Shanghai-based optical computing company Guangbenwei (Optical Computing Tech) has partnered with Dongfang Tianxuan to develop space-based photonic computing payloads, using photons instead of electrons to bypass radiation, heat dissipation, and power challenges in orbit.
Google informs Meta that it will limit access to its Gemini AI due to capacity constraints.
OpenAI has previewed GPT-5.6 models and announced new pricing.
Chinese AI company Z.ai publicly claims its cybersecurity capabilities are on par with Mythos, according to a report by The Verge.
Baidu has released a new OCR model capable of processing up to 40 pages in a single pass.
The 14th Internet Security Conference opens in Beijing, focusing on moving from consensus to action in security governance in the age of agents.
xAI unveils Grok 4.5, a significant upgrade powered by a 1.5 trillion parameter V9 model.
Ford admits it mistakenly thought that just introducing AI would produce high quality products, and is now rehiring experienced engineers to bridge the gap.
The US government has lifted the ban on Anthropic's Claude Fable 5 AI model, which was previously restricted due to security concerns.
Baidu's AI chip subsidiary Kunlunxin is reportedly planning a Hong Kong IPO with a target valuation of up to $50 billion, according to The Economic Times.
A vice president from Apple's Vision Pro team has left to join OpenAI.
Google has placed limits on Meta's use of its Gemini AI models, according to a Financial Times report. The specific terms have not been disclosed.
The Create2026 AI Developer Conference has grandly opened with the theme 'All Things United, Self-Evolution,' exploring the future of the AI agent era.
An Indian startup has filed a brand dispute with the SEC against Anthropic ahead of its IPO.
Hackaday covers a project that teaches an AI to play a racing game using only raw screen pixels as input, showcasing a vision-based approach to game AI training.
HP Inc. has entered a Frontier strategic partnership with OpenAI, planning to deploy AI across customer experience, software development, and enterprise operations.
A paper submitted to the ACL conference demonstrates an AI system that constructs complete artificial languages from scratch, outperforming all previous conlang generation systems across every metric.
Investors hunting for the next big AI winner after Nvidia are turning their attention to US memory chip maker Micron Technology.
Amazon has pulled the plug on Luca Guadagnino's OpenAI-themed film 'Artificial.' The director says he is 'not surprised' by the decision.
Australia's Firmus Technologies has signed an AI access deal with Nvidia, as reported by Reuters.
Google has restricted Meta's access to Gemini due to a surge in compute demand, according to reports.
Anthropic claims that Alibaba may have replicated the capabilities of its Claude model simply by asking questions.
Okta has become the first company to implement AI agent governance within FedRAMP boundaries, as reported by The New Stack.
Austria is pushing the European Union to host Anthropic AI as a response to tightening US AI regulations, according to a Bloomberg News report.
Austria has submitted a bid to become the host location for Anthropic's AI operations in the European Union.
Zhipu AI has claimed a breakthrough in advanced vulnerability detection technology, as reported by HOKANEWS.COM.
DeepSeek has officially published the Eagle3_Gemma4_12B_TTT7 model on Hugging Face, built on the Gemma 4 architecture with TTT7 technology.
DeepSeek has officially published the Eagle3_Qwen3_14B_TTT7 model on Hugging Face, built on the Qwen3 14B architecture with TTT7 technology.
DeepSeek has uploaded the new Eagle3 model eagle3_qwen3_8b_ttt7, based on Qwen3 architecture, to its official Hugging Face repository in safetensors format.
Lanzhou University has officially established a School of Artificial Intelligence, further expanding AI education and research in Northwest China.
According to Sina Finance, 12 AI models made only 5 correct predictions out of 24 tasks, for an accuracy rate below 21%.
Google has restricted Meta's access to its Gemini AI models, according to NewsBytes.
DeepSeek claims its DSpark technology accelerates AI processing and reduces costs while alleviating computational bottlenecks and chip supply pressure.
OpenAI was forced to issue a limited release of GPT 5.6 following White House pressure over safety concerns, marking a new chapter in U.S. government-AI regulatory dynamics.
According to a report from Dawn, Google has restricted Meta's use of its Gemini AI models.
Google has restricted Meta's access to its Gemini AI models, with other Google Cloud clients also impacted, according to a report from Inshorts.
Alibaba Cloud has been recognized as a leader in Omdia's 2026 Market Radar report for Agentic AI Cloud services in Asia and Oceania.
WinBuzzer reports that the release of OpenAI's GPT-5.6 has encountered a government approval hurdle.
Daily Archive